HfflBnr]
■ i
r :»i i ■. i
It
■ r i
Ag ide to lines of the French where White ets the pace
Steffen Pedersen

French: Advance
and Other Lines
Steffen Pedersen
A ED BUT

First published in the UK by Gambit Publications Ltd 2005
Copyright © Gambit Publications 2005
The right of Steffen Pedersen to be identified as the author of this work has been asserted in
accordance with the Copyright, Designs and Patents Act 1988.
All rights reserved. This book is sold subject to the condition that it shall not, by way of trade or
otherwise, be lent, re-sold, hired out or otherwise circulated in any form of binding or cover other
than that in which it is published and without a similar condition including this condition being
imposed on the subsequent purchaser.
A copy of the British Library Cataloguing in Publication data is available from the British Library.
ISBN 1 904600 40 9
Distribution:
Worldwide (except USA): Central Books Ltd, 99 Wallis Rd, London E9 5LN.
Tel +44 (0)20 8986 4854 Fax +44 (0)20 8533 5821. E-mail: orders@Centralbooks.com
USA: Continental Enterprises Group, Inc., 302 West North 2nd Street, Seneca, SC 29678, USA.
For all other enquiries (including a full list of all Gambit chess titles) please contact the publishers,
Gambit Publications Ltd, 6 Bradmore Park Rd, Hammersmith, London W6 ODS, England.
E-mail: info@gambitbooks.com
Or visit the GAMBIT web site at http://www.gambitbooks.com
Edited by Graham Burgess
Typeset by Petra Nunn
Cover image by Wolff Morrow
Printed in Great Britain by The Cromwell Press, Trowbridge, Wilts.
10 987654321
Gambit Publications Ltd
Managing Director: GM Murray Chandler
Chess Director: GM John Nunn
Editorial Director: FM Graham Burgess
German Editor: WFM Petra Nunn

Contents
Symbols 4
Bibliography 5
Introduction 6
The Advance Variation
1 Advance: S...!^ 6 a3 13
2 6 ±e2 42
3 6 Jtd3 and the Milner-Barry Gambit 48
4 5...i.d7 54
5 5...£>ge7 and 5...£>h6 84
6 Various Deviations 91
Exchange Variation, King's Indian Attack Set-Ups and Rare Lines
7 Exchange Variation 102
8 King's Indian Attack 112
9 Wing Gambit and Other Rare Lines 135
Index of Variations
142

Symbols
+
++
#
II
1
!?
?!
?
??
+-
±
±
=
oo
+11
+
-+
Ch
Cht
Wch
Wcht
Ech
Echt
ECC
Ct
IZ
Z
OL
jr
worn
rpd
tt
sim
corr.
qual
1-0
V2-V2
0-1
(")
(D)
check
double check
checkmate
brilliant move
good move
interesting move
dubious move
bad move
blunder
White is winning
White is much better
White is slightly better
equal position
unclear position
Black is slightly better
Black is much better
Black is winning
championship
team championship
world championship
world team championship
European championship
European team championship
European Clubs Cup
candidates event
interzonal event
zonal event
olympiad
junior event
women's event
rapidplay game
team tournament
game from simultaneous display
correspondence game
qualifying event
the game ends in a win for White
the game ends in a draw
the game ends in a win for Black
nth match game
see next diagram

Bibliography
Books
ECO, volume C (2nd, 3rd and 4th editions), Sahovski Informator, 1981/1997/2000
Tiemann: Die Franzosisch Verteidigung (band 2), Reinhold Dreier, 1992
Uhlmann: Winning with the French, Batsford, 1995
Watson: Play the French (new edition), Cadogan, 1996
McDonald and Harley: Mastering the French, Batsford, 1997
Harding: Four Gambits to beat the French, Chess Digest, 1998
Dunnington: The Ultimate King's Indian Attack, Batsford, 1998
Kosten: The French Advance, Chess Press, 1998
Minev: French Defense 2, Thinkers' Press, 1998
Nunn, Burgess, Emms, Gallagher: Nunn's Chess Openings (NCO), Gambit/Everyman, 1999
Pedersen: The Main Line French: 3 Q±c3, Gambit, 2001
Watson: Play the French (3rd edition), Everyman, 2003
Psakhis: Advance and other Anti-French Variations, Batsford, 2003
Periodicals
Various, including Informator, New In Chess Yearbook and ChessBase Magazine.

Introduction
This book covers a rather broad variety of
variations against the French. In this introduction I
shall provide a brief overview here of the
following systems:
• The Advance Variation
• The Exchange Variation
• The King's Indian Attack
• The Wing Gambit
The Advance Variation is clearly the largest
subject of the four and the often rather messy
lines have in recent years attracted many
attacking players, such as Shirov and Grishchuk
to name just two.
The Exchange Variation is a different matter
altogether. It often leads to very symmetrical
play, and can lead to a quick draw if both
players are in the mood. However, it is by no means
a forced draw, and Black can take comfort in
the fact that he has fewer problems equalizing
than in most opening lines.
The King's Indian Attack is often chosen by
white players as a 'system', where the first 8-10
moves can be played very schematically and
thus little thought is required. However, it is also
a system where there is room for creativity, and
interesting unbalanced play generally results.
The Wing Gambit is an aggressive choice and
shouldn't be underestimated. However, while it
often seems to give White a good deal of play in
practice, we can also draw a relevant conclusion
about its soundness from the fact that it is very
rarely chosen by high-level players.
The Advance Variation
With the Advance Variation (see following
diagram) White immediately limits Black's choice
of variations, which is otherwise (against 3 £\d2
and 3 £\c3) fairly broad. With the advance of the
e-pawn, White grabs space and hinders Black's
B
Wffifc""' W
■ in*
AH "
mXmwmmm
i i
flff
ififtiH
lAH
wm
kingside development. As a consequence of
White's space advantage, Black often has
trouble finding useful positions for his minor pieces,
and if he isn't careful they will end up away
from the kingside with no chance of providing
their king with sufficient protection. It follows
that White's chances often lie on the kingside
and a sacrificial attack is not at all uncommon.
Here we see such an instance in one of the
earliest known games in the Advance Variation:
L. Paulsen - Schwarz
Leipzig (3) 1879

Introduction
7
White already has a nice attacking position.
Probably wanting to avoid Jtxf5 followed by
Sxd5 without weakening his queenside further,
Black ran into the thematic Greek Gift
sacrifice...
15...£>fe7?
It was necessary to play 15...b6 followed by
...Wd8.
16 ±xh7+! *xh7 17 £>g5+ *g6
Other moves are quickly refuted:
a) 17.„*g8 18 #h5 Sfe8 19 #xf7+ <4>h8
20 Sc3 is soon mate.
b) 17...*h6 18 #g4 Sh8 19 Sc3, etc.
18 #g4 f5 19 #g3! (D)
19...#c8?!
Black should try 19...f4 but 20 #h4 Sh8 21
#\f4 Iaf8 22 #g4 keeps the white attack go-
ins.
20 lc3
Much better than 20 £>xe6+ <4>f7 21 #xg7+?!
s\e6. after which White has nothing convinc-
20...f4 21 #g4 5M5 22 Sh3 Sh8 23 £>xe6+
Sf7 24#xf5+*e7
24...*g8 25 Sxh8+ <4>xh8 26 #h5+ <4>g8 27
-lp i.f5 28 #f7+ <4>h8 29 #xd5 +-.
25 #g5+! *xe6 26 #g6+ <4>e7 27 #xg7+
1-0
A typical way for Black to arrange his minor
r eces on the kingside is to move the knight to
5 and put the bishop on e7. However, the
eight's position on f5 is far from safe and may
rji to two types of problems:
1) White can exchange his bishop for it,
thereby doubling Black's pawns on the f-file,
isolating one on d5, and in general leaving
Black with a very bad bishop;
2) White may simply push it away with g4.
The former problem is in general less
threatening since Black's position remains solid if
undynamic. In the following example Black ran
into trouble after White pushed the knight away
with g4:
Hodgson - Lalic
Port Erin 1995
Black's last move, 13...0-0, is queried by
Hodgson, who suggests 13...h5 instead.
14 g4 £>h4
Here Black should seek not at all untypical
complications with 14...£>h6 15 £>e3 f6!?.
15 £ixh4 ±xh4 16 f4 f6?
16... Jte7 is necessary but White has a
promising attack after the precautionary 17 ^hl.
17 g5! fxg5
On 17...fxe5, White can play 18 ±xh7+
<4>xh7 19 #h5+ <4>g8 20 g6, but the simplest
might be 18 #h5 e4 19 J.e2, winning a piece.
18 #h5 h6
18...g6 19 ±xg6 hxg6 20 #xg6+ <4>h8 21
fxg5 #d8 22 #h6+ <4>g8 23 Sf6 +-.
19 fxg5
19 #g6 Sf5 20 Axf5 exf5 21 4^e3 also wins.
19...±xg5 20 #g6 Sf5 21 ±xf5 exf5 22 h4!
±xh4 23 £>e3 #d8 24 £>xd5 ±e8 25 #g2
*h7 26 £>f4
White is winning.

8
FRENCH: ADVANCE AND OTHER LINES
The next example shows an interesting twist
as Black's knight is sidelined on h6 early in the
game, White manoeuvres on the queenside to
improve his coordination and avoid counter-
play, and then eventually decides things on the
kingside.
I
■ lo HI
ff fl ^gp A%
right. 17...#d8 is better, but White keeps an
advantage with 18 #e3 f6 19 exf6 Sxf6 20 £>e5.
18 £>xd4 #xd4 19 £>e4 #b6 20 £>f6+!
±xf6 21 exf6 ±b5 22 ±xb5 #xb5 23 #e3 d4
24±xd4 2fd8(D)
Timman - Cu. Hansen
Malmo/Copenhagen (2) 2003
12...Sc8
Black doesn't want to castle yet and
improves his position by placing the rook on the
open c-file. By delaying castling, Black
prevents £fc3 for the time being (this would be the
reply to 12...0-0 since Black cannot then
capture on d4).
13 £>bd2
With this Timman abandons the idea of
getting the chance to develop the knight on c3.
That idea could have been pursued by 13 #d2
but this would result in another problem for
White, namely that the f3-knight is now
unprotected, and thus Black can play 13...0-0 14 £ic3
f6!.
13...£>b8
Preparing... Jtb5.
14#e2a615£>bl!(T>)
The very slight change in the position is
enough to convince Timman that he can direct
the knight to c3 all the same.
15...£>c6 16 0-0 0-0 17 £>c3 £>xd4?
A wrong decision which isn't immediately
obvious but Timman's feeling that it would be
too dangerous for Black to take the pawn was
W
I
ZmmM?. y/y
H H * m, *
A IP IP A ill ^
Wiliiriif IP IP
IIWi§ wm wk
M W.
25 Sfdl?!
White is still winning after this but Black is
without counterplay after 25 fiacl!.
25...Hd5 26 i.b2 Sg5 27 Had He8 28 ff4
Sg6 29 Sc5 «e2 30 Sd2 fel+ 31 *h2 I'll 32
Sh5 4M5 33 «f3 £>h6 34 fxg7 f5 35 g5 £if7 36
£f6
36fixh7!+-.
36...£>xg5?! 37 ±xg5 h6 38 fidl #b5 39
i.xh6 ®o5+ 40 Shi Wf6 41 4*3 1-0
The g4 advance isn't always attractive, and
sometimes Black allows it as part of a strategy
of provoking weaknesses. Quite often this also
turns out successfully:

Introduction
9
Alekseev - Se. Ivanov
St Petersburg 2001
le4e62d4d5 3e5c5 4c3Wb6 5&f3&c66
a3 £>h6 7 M cxd4 8 cxd4 £>f5 9 £b2 £d7 10
g4£rfe7
In Timman-Cu.Hansen from above, Black
had gone 10...£>h6.
Il£te3&a512&d2
A month later, in a game between the same
two players, White chose 12 #c2!?.
12...Sc813Scl£g6!(D)
It is a problem with 10...£tfe7 that White is
allowed to develop his knight on c3 but Black's
knight also has better prospects from here.
14£b5
To justify the weaknesses created by White's
aggressive pawn moves, he must keep up the
tempo and not allow Black to finish his
development smoothly. The main alternative is 14
h4, but for more details on the theory of this
line, you should consult Line C21 of Chapter 1.
14...£>c6 15 £>f3 £e7 16 h4?!
This may be taking the aggressiveness a little
too far but also after the more cautious 16 JLe2
0-0, intending ...f6, Black has a good game.
16...£>cxe5!
A classical little combination. This was not
overlooked by White but he doesn't get as much
play for the pawn as he would have liked.
17 dxe5 £.xb518 h5 £tf819 £>xb5 Sxcl 20
#xcl #xb5 21 #c8+ £d8 22 <&d2 £>d7 23
£>d4#a6
Black is winning.
We have already seen from above that White
may advance his queenside pawns with a3 and
b4. This is a modern treatment of the Advance
Variation and may be seen as a combination of
aggression and restriction. It is not least a
preventive policy, aiming to keep Black's pieces
back and getting b2 for the dark-squared bishop
to defend d4. But ideally it may also serve as a
support for the manoeuvre £fol-c3-a4-c5. The
drawback to all this is that Black gets
something to attack, for example with ...a5, and that
White weakens the light squares. The following
is a good example of how badly things can also
turn out.
Alekseev - Rustemov
Russian Cht 2001
le4e62d4d53e5c5 4c3Wb6 5£>f3i.d7 6
a3 i.b5 7 b4
7 c4!? is the sharper approach and the reason
that Black sometimes chooses to exchange on
d4 before playing .. JLb5.
7...cxd4 8 £xb5+ #xb5 9 cxd4 £>d7!
Already with an eye on c4.
10 £>c3 #c6 11 £b2 &b6 12 £>d2 &e7 13
0-0£>f514Scl#d7(T>)
The position is already pleasant for Black.
Whether it is objectively better for him is
difficult to say but the long-term problem of the bad
bishop was now enough to provoke White into
weakening his kingside.
15£>e2i.e716g4?!
16 £\g3!? is the natural alternative.

10
FRENCH: ADVANCE AND OTHER LINES
16...£>h4 17 £>g3 0-0 18 f4 f5 19 exf6?
19 g5 would of course only have made the
problem with the dark-squared bishop worse,
but should have been preferred anyway.
19...Ixf6 20 We2 Iaf8 21 Icel Ixf4 22
fixf4 Ixf4 23 £>h5 If7 24 Wxe6 ±g5!
There is a problem with a fork on f3.
25 We8+ Sf8 26 Wxd7 £>xd7 27 Acl £xd2
28 Ie7 If 7 29 Se8+ £>f8 30 &xd2 £>f3+ 0-1
In general many of the early moves in the
Advance Variation form part of a major battle
for d4.
Black can attack the d4-pawn naturally with
the moves ...£>c6, ...Wb6 and a knight to f5. The
route may be via e7 (often preceded by an
exchange on d4 in order to avoid dxc5), or via h6
(which sometimes allows Jtxh6). In turn White
can defend d4 with the obvious £if3 but a further
defender is needed, and it is not so obvious what
piece it should be. Often White plays £&3-c2, or
sometimes Ae3 (of course only when Black
cannot just take on b2). Playing the knight to c2
is often done in lines where Black goes ...£}ge7,
so there is no ... Jbta3 opportunity. In these lines
Black can often try to exchange off the c2-knight
with ...^b4, swapping an attacker for a defender
but in some positions leaving White with a
problem since he cannot allow Black to make the
capture on c2 as this would force the queen away
from the defence of d4. In all this we must
remember that Black will often be the one with the
lead in development and checks on the a5-el
diagonal could turn out annoyingly for White.
This leads us to the a3 + b4 advance, which
prevents these problems on the dark squares on
the queenside and provides White with an extra
option of how to defend d4, namely with Jtb2.
Apart from Black being able to attack b4
with ...a5, and potential problems on the light
squares that we have come across above, we
should discuss a further type of position. That is
when White plays an early a3 and Black closes
the game with ...c4. The following position is
then very common:
This could also be with White's bishop on e2
instead of the move g3. Strategically this is very
interesting as play is divided strictly between
Black playing on the queenside and White on
the kingside. Black's plan is laid out
graphically with good control of the light squares on
the queenside and appealing manoeuvring
possibilities. In the long run this might even result
in a chance to make real progress by advancing
the a- and b-pawns, eventually creating a target
in White's position. Meanwhile White would
of course advance on the kingside.
The Exchange Variation
The exchange of pawns in the centre makes the
position very drawish if both sides continue in
stereotyped fashion. It is easy to imagine (and
this often happen) that both sides develop with
something such as £>f3/...£>f6, i.d3/...i.d6,
Jtg5/...Jtg4 and then all the major pieces
disappear on the open e-file.

Introduction
11
However, things don't need to be as
unimaginative as that, and both White and Black have
options to avoid symmetry if they want.
With 4 c4, for example, White can strive for
a position with an isolated d-pawn and active
piece-play. And against 4 £jf 3 or 4 Jtd3 Black
can choose a set-up with ...£jc6 and sometimes
even aim to castle queenside.
The King's Indian Attack
For many players the King's Indian Attack is a
good practical choice, not only against the
French but also because it can be played against
the Sicilian and even the Caro-Kann as well.
White may play the first 6-10 moves (or even
more) basically without thinking and according
10 a specific scheme which gives Black few
opportunities to interrupt the standard way of de-
\ eloping. However, with a King's Indian Attack
set-up there is also room for flexibility and
creativity.
In the presentation of the King's Indian At-
.dck I have adopted a sort of repertoire-based
: overage, restricting my coverage to positions
:n which Black fianchettoes his dark-squared
?ishop.
The following diagram shows one of the
-:andard positions. If Black castles kingside,
•Vhite's idea would be to close the centre by
rlaying e5 and then manoeuvre the knight
iround with, for example, £jfl, h4, £jh2-g4,
.: to support the e5-pawn with c3 and d4. This
type of plan is a little annoying for Black, who
will face difficulties on the kingside as the e5-
pawn hinders support from Black's pieces on
that side.
Hence we will often see a kind of shadow-
boxing as Black will not castle yet, and White
will also only play e5 if Black has castled.
Therefore in the above position Black's most
common move is ...b6. Then depending on
White's choice Black may play further semi-
waiting moves such as ...a5, ...Jta6 and ...Wc7.
If White plays e5 before Black has castled,
then the move ...g5 is often strong, thereby
indirectly attacking the e-pawn as Black threatens
...g4. Even in some positions where Black has
castled, this idea should be taken into
consideration. Here is a well-known game that illustrates
the difficulties Black may face on the kingside:
Ciocaltea - llijin
Romanian Ch (Timisoara) 1976
Ie4c5
Although this game starts as a Sicilian, it
quickly transposes into lines that are relevant to
the French KIA.
2 £>f3 e6 3 d3 d5 4 £jbd2 £>c6 5 g3 g6 6
±g2 ±g7 7 0-0 £>ge7 8 Sel 0-0 9 e5
9 h4! h6 10 e5 is another option, and rules
out Black's 10th move alternative.
9...«c710«e2b6
10...g5! has scored well for Black.
11 £jfl ±a6 12 .4.14 2ad8 13 h4 d4
Black's last few moves doesn't look right,
but already the type of position is much easier

12
French: Advance and Other Lines
to play for White. Now the attack develops
instructively.
14 £>lh2 £>b415 *d2! Sfe8 16 £>g4 £>ed5
17 ±h6 ±h8
It is clear that Black's position cannot bear
an exchange of the dark-squared bishops.
18 ±g5 Sd7 19 a3 £>c6 (D)
20 &f6! £>xf6 21 exf6 h5 22 £>h6+ <&f8 23
£>g5 £xf6 24 £>h7+ <&>g7 25 £>xf6 *xf6 26
£>f5!
Black hasn't defended badly but this is the
culmination of a powerful attack. Black is
helpless as captures other than the text-move allow
27 *g5#.
26...exf5 27 Sxe8 *g7 28 Sael £>d8 29
*g5c4 30Hh8!
A nice finish to a game where the main
theme throughout was White's play on the dark
squares.
30...£>e6 31 *h6+ <&f6 32 Se8 1-0
The Wing Gambit
One of the more interesting gambits against the
French is the Wing Gambit, where White
sacrifices a pawn to win control of the centre. I am
sure that Black has no problems theoretically
but in practice White's initiative should not be
underestimated.
A main line runs 1 e4 e6 2 £}f 3 d5 3 e5 c5 4
b4 cxb4 5 a3 £k6 6 axb4 ±xb4 7 c3 ±e7 8 d4
If Black plays passively, White develops a
powerful attack as after Jtd3 White has both his
bishops aimed menacingly towards the black
kingside. As so often in the French, an attack
against White's centre with 8...f6 is regarded as
the strongest, after which the knight is
developed via h6 even at the cost of having pawns
doubled on the h-file.

1 Advance: 5...Wb6 6 a3
le4e6 2d4d5 3e5c5 4c3£>c6 5£>f3lT>6 6
2&(D)
P
P^J«
P ■ PAH
White's 6th move intends queenside
expansion with b4. It further neglects minor-piece
development, and thus seems time-consuming,
but it has certain advantages as Black will find
himself in an almost claustrophobic situation if
White can maintain his pawn-chains. Besides,
simply guarding the b4-square is often useful.
And, just as important, once White has
advanced his b-pawn, he can develop his dark-
squared bishop, which is otherwise assigned to
protecting the b2-pawn.
A very logical reply is 6...c4, closing the
queenside and making a b4 advance less
attractive, and aiming to exploit the weakness on b3.
This has always been a popular reply but also
has its drawbacks because gaining something
from the single weakness on b3 requires lots of
footwork and meanwhile White enjoys free play
on the kingside, due to his space advantage on
that side.
Thus, a number of other ideas have been
worked out by Black, by which he allows White
to gain more space but aims for a quick
counterattack. These ideas are often a blend of attacks
against White's centre and/or the queenside.
Having made many pawn moves, White can
often find himself slightly lacking development.
A: 6...f6!? 13
B: 6...a5 14
C: 6...£>h6 16
D: 6...c4 28
E: 6...±d7 36
A)
6...f6!? (D)
While a major option in many similar lines,
this hasn't been very popular in this exact
position. The idea is much more common in the line
5...k.d7 6 a3, and now 6.../6, where the attack
against e5 gains in strength with the option of
following up with ..Mel and ...0-0-0.
W
7±d3
7 b4!? is also an option and, in view of
Murei's idea of playing 8...£Mi6 in our main
line, this might even be preferable:
a) After 7...cxd4 8 cxd4 fxe5 9 dxe5 a5 10
b5 Black went for the speculative 10... Jtc5?! in
Grujic-B.Maksimovic, Niska Banja 1996, but
White can simply play 11 bxc6! ±xf2+ 12
4>d2 (12 *e2 bxc6 13 Wc2, intending *dl ±)
12...*e3+ 13 4>c2 Wc5+ 14 4>b2 bxc6 15 4>a2

14
FRENCH: ADVANCE AND OTHER LINES
±. 10...£>d8, intending ...g6, ...£tf7, ...£te7 and
...^.g7, is more circumspect.
b) Watson draws attention to 7...fxe5!? 8
dxe5 £>h6!, when he claims 9 ^.xh6 gxh6 10
Ad3 Ag7 11 0-0 0-0 (Guigonis-Renaudin,
Paris 2000) to be slightly better for Black.
Certainly, Black should at least be OK in this.
Probably White should play something like 9 b5
£>a5/£>e7 10 c4.
7...fxe5 (D)
Watson likes 7...c4 8 Ac2 g6!?. His main
line runs 9 0-0 fxe5 10 £>xe5 £>xe5 11 dxe5
£>h6, continuing 12 We2 Ag7 13 Ae3 Wc7 14
f4 0-0 with a balanced position. I tend to agree
that Black is alright when play develops
quietly, although I am not sure whether I wouldn't
rather choose to be White at the end of this.
Moreover, White has several other ways of
playing. 12 JLa4+ springs to mind, when 12...JLd7
13 &xd7+ 4>xd7 14 £>d2 followed by b3 could
be a problem for Black with his king in the
centre, while 12...*f7 13 Wf3+ 4>g8 14 £>d2,
aiming to bring the knight to f3, also looks better
for White.
8dxe5
In many similar lines it is advisable to take
with the knight on e5 first, but here the text-
move is the most common. 8 £>xe5! £>xe5 9
dxe5 g6 is solid for Black, but with best play
White should keep an edge:
a) 10 c4 Ag7 11 We2 £>e7 gives Black no
problems.
b) 10h4c411ix2ix5 12lfe2£>e7 13h5
Sf8!? 14 0-0 Ad7, intending ...0-0-0 offers
Black slightly better chances, F.Roder-Kugel-
mann, Marktredwitz 1984.
c) 10 0-0! Ag7 11 We2 £>e7 12 *hl c4 13
Ac2 0-0 14 f4 Ad7 15 £>d2 £>f5 16 «M3 with
slightly better chances for White.
8...£>h6!?
This is Murei's idea, which has turned out
surprisingly well for Black.
9 0-0
9 c4 £tf7 10 £\c3 d4 doesn't look like a
problem for Black.
9...c4 10 i.c2 £>f 7 (D)
11 b3
In Afek-Murei, St Quentin 1999 Black had
reason to be satisfied with the opening after 11
We2 g6 12 Sa2 Ag7 13 i.f4 0-0 14 £>bd2 Wc7
15 Sel £>h6 16 g3 b5.
Il...cxb3 12 &xb3 Ae7 13 Sa2 &d7 14
Se2 £>a5 15 Ac2 &b5 16 &a4 0-0
Black stands slightly better, Dvoirys-Murei,
Dieren 2000.
B)
6...a5 (D)
The most radical way of preventing a white
advance on the queenside. Many would argue
that it seriously weakens the b5-square. True,
but White has also weakened his queenside,
and it is only logical that Black wants to grab
some space on the side where it is possible. On
the other hand, there are indeed some lines
where the weakening of b5 could be fatal.
7i.d3!

ADVANCE: 5..Mb6 6 a3
15
w
Emm Jl. mm _A_ m%fc Jfei ^sK
mJLm+mbM
^ABl
mm Hill i
This is the most serious test of Black's last
move. White plays a Milner-Barry Gambit (see
Chapter 3) with the two a-pawn moves
interposed. Clearly, this is to White's advantage, but
will it be significant enough to promise him a
real advantage? The alternative is 7 JLe2 cxd4 8
cxd4 £>ge7 9 £>c3 £rf5 10 £>b5 h5 with roughly
equal chances.
7...±d7 (D)
It is very useful in the following moves to
compare with the Milner-Barry Gambit
(henceforth MBG). In the real MBG, this is
considered inaccurate in view of dxc5, but now White
has wasted a move on a3, which would
probably make this capture less strong. If Black
instead adopts the prescribed MBG move-order
with 7...cxd4 8 cxd4 JLd7 White could consider
keeping his pawn with 9 JLc2, when a3 looks a
lot more useful than ...a5.
8 Jtc2 is solid but not really in the spirit of the
gambit. It must be said, though, that it is better
than in the real MBG since Black doesn't have
the option of harassing the bishop with ...^b4.
Black has several ambitious approaches then,
including ...f6, but Korchnoi has on several
occasions just gone 8...h5, intending to develop by
...£ih6 and .. Jte7, with a solid position. An
example is Mukhametov-Korchnoi, Baden 1997:9
0-0 £>h6 10 b3 i.e7 11 h4 cxd4 12 cxd4 fic8 13
Ab2 #d8 14 g3 b5 15 £>bd2 Wb6 16 lei b4 17
a4 4>d8!? 18 £>f 1 4>c7 19 £>e3 4>b8 20 Icl flc7
21 Wd2.fi.c8 with equality.
8...cxd4 9 cxd4 £>xd4 10 £>xd4 Wxd4 11
£>c3
A largely ignored possibility in this precise
position is 11 We2 (D), which is a little curious
since this has become the preferred line of
some MBG specialists (without a3 and ...a5).
Black can then try:
I
u^^i
A mm A mm x vm? x
mm a mm mm
*ww mm
m
mm
4«
*J
y mm^ mz^ mm, w
B ■&■ ■ j
if WVM'M A ¥?t
\v
8 0-0
a) ll...#b6 12 £)c3 ±c5 13 Ad2 a4 14
£ib5 £te7 15 Sacl i.xb5 16 £.xb5+ £>c6 17
2xc5!? WXC5 18 ±b4 #b6 19 Bel with very
good compensation for White, Schorra-Rag-
narsson, corr. 1997.
b) Il...£>e7 12£ic3trb6(12...£ic613.£.e3
fxeS 14 f41U6 15 £)b5 Wb8 16 f5 e5 17 #h5
looks dangerous for Black; in the MBG Black
has the option of ...a6 to kick the knight away)
13 i.e3 d4 14 £)b5 £rf5 15 £ixd4! £>xd4 16
Wg4 £x5 (16...Wxb2 17 ±xd4 is at least
compensation for White) 17 b4! and White is better.
c) ll...f6!? 12exf6(12Wh5+*d8 13tn
£ie7 is very good for Black but 12 J.b5!?, as in

16
French: Advance and Other Lines
Harrouin-Savoroux, corr. 1993, is worth
considering) 12...£>xf6 13 £>c3 i.d6 14 i.e3 (14
£>b5 We5 15 £>xd6+ «xd6 16 ±d2 0-0 +)
14...Wh4 15 g3 Wh3 16 £>b5 ±xb5 17 ±xb5+
&f7 and Black is clearly better.
We now return to 11 £>c3 (D):
ll...£>e7!
This idea has not been seen in the MBG in
exactly this form. Even though the idea in itself
is very common, Black would usually take the
precaution of playing ...a6 first. That's illegal
here but there are a few other moves that
deserve attention:
a) ll...»xe5 12 Sel «d6 (12...Wb8 is
recommended in the MBG but here White has 13
£>xd5 ±d6 14 £>b6! ±) 13 £>b5 Wb8 (13...±xb5
14 ±xb5+ &d8 15 ±e3 ±) 14 Wf3 ±d6 15
£>xd6+! «xd6 16 ±f4 We7 17 Wg3 with very
good compensation for White, Nemec-Vrbata,
Czech Cht 1998.
b) ll...Wb6 12 Wg4 (in Benderac-Vukic,
Bosnian Cht 2003 White was successful with 12
£>b5!? ±xb5 13 Wb3, but this requires further
investigation) 12...g6 13 Jte3 Jtc5 was played
in Shirov-Anand, Teheran FIDE Wch (1) 2000.
Even though Anand drew the game, the general
impression was that White was better after the
opening. Ftacnik gives 14 Wf4!, and claims an
advantage for White in all lines below:
bl) 14...£>e7 15 «f6 Sg8 16 ±xc5 «xc5
17 Sacl Wb6 18 Sc2! (18 a4 Sd8 19 £>b5
±xb5 20 i.xb5+ &f8).
b2) 14...d4 15 b4!! dxe3 16 bxc5 Wxc5 17
£>e4 +-.
b3) 14...Sc8 15 £>a4! ±xa4 (15...±xe3 16
fxe3 ±xa4 17 «xa4+ &f8 18 Hxf7+! &xf7 19
Wd7+ £>e7 20 Sf 1+ wins for White) 16 Wxa4+
&d8 (16...&f8 17 Sacl! +-) 17 ±xc5 Sxc5 18
b4±
12£>b5
A slower approach like 12 Sel ^c6 13 ^b5
Wb6 14 &e3 Wd8 15 ±f4 (Cools-Claesen,
Antwerp 1997) might also be feasible.
12...Wxe5 13 Sel «b8 14 «f3 ±xb5 15
±xb5+ £>c6 16 «xd5 «d6! 17 «f3
White has very slight pressure but
objectively I think Black is doing OK, S.Pedersen-
L.B.Hansen, Danish Ch (Aalborg) 2000.
C)
6...£>h6 (D)
7b4
This is White's most consistent
continuation.
7 ±xh6? «xb2 8 ±cl Wxal 9 «c2 c4!,
intending ...£>a5 and ...Jtd7, is good for Black.
Black's queen is stuck in the corner but White
cannot trap it, and cannot exploit it in other
ways since the queen gets out as soon as White
tries to develop his queenside pieces.
7 Jtd3 is feasible though, but Black can be
satisfied with 7...cxd4 8 cxd4 £>f5 9 ±xf5 exf5
10 £ic3 Jte6. This position is briefly dealt with
via the move-order 5...&d7 6 a3 Zhge7 7 &d3
cxd4 8 cxd4 ^b6 9 kcl *hf5 10 kxf5 exfS 11
£hc3&e6=.
7...cxd4 8 cxd4

ADVANCE: 5..Mb6 6 a3
17
8 ±xh6 gxh6 9 cxd4 (D) is White's other
option, and a common theme in various ...£\h6
lines, but White is probably aiming for too
much with this. It is true, Black's kingside is
weakened but Black has an important
advantage in the bishop-pair and a possibility of
opening the position with ...Jtg7 and ...f6.
Now:
a) 9..AgJ 10£>c3 f6 (10...0-0 11 £>a4Wd8
12 Jtb5 gives White a slight advantage) 11 £\a4
#c7 12 i.b5 fxe5 13 i.xc6+! followed by 14
£\xe5 is very good for White.
b) 9...Bg8?! 10 £>c3 (10 h3 to prevent ...Bg4
seems very slow) 10...2g4 (10...£kb4 might be
a dangerous sacrifice, although I doubt it to be
correct even after 11 axb4 Jtxb4, and White
might do better still with 11 £\a4 Wa5 12 axb4
jLxb4+ 13 4^2, when the king is going to be
fairly safe on e3) 11 h3 2f4 12 i.b5 a5 13 0-0!
axb4 14 axb4 2xal 15 Wxal 2xf3 (15...Axb4
16 ±xc6+ bxc6 17 £>e2 2xf3 18 gxf3 ±a6 19
Bbl! +-) 16 gxf3 «xd4 17 Wa8 and White
was winning in Keitlinghaus-Blauert, Budapest
1998.
c) 9...Ad7! 10 ±e2 (10 ^c3?! £>xb4! 11
axb4 &xb4 12 Wb3 2c8 13 Bel Wa5 14 <4>d2
0-0 looks too dangerous for White) 10...Bc8 11
•3-0 Ag7 12 Wd2 (D) (other moves make less
sense; 12 Wd3 is best met by 12...f6!, when 13
exf6 JLxf6 is good for Black).
Now:
c 1) 12...£kd4 is given by Watson but I don't
quite trust it. After 13 £>xd4! (better than 13
Sxd4 Wxd4 14 £>xd4 ±xe5 15 £>b3 Axal 16
£>xal ±a4 17 £>d2 <4>e7 + Watson) 13...Axe5
he continues 14 Bdl i.a4 15 We3 ixdl 16
«xe5 (16 ±b5+? «xb5 17 «xe5? «d3! 18
Wxh8+ <4>d7 19 We5 ±a4 -+) 16...Axe2 17
£\xe2 0-0, which he assesses as slightly better
for Black. I wouldn't be so positive about
Black's prospects, although I admit that White's
task of mobilizing his knights after something
like 18 £>bc3 f6!? 19 Wh5 «a6 20 «xh6 Bc4
isn't easy. One also has to examine 14 £ft>3
i.a4 15 £>c5 ixal 16 £>xa4 Wd4.
c2) 12...0-0 13 Ba2!? (13 ^c3?! £>xd4! 14
£>xd4 ±xe5 15 Badl ±xd4 16 £>xd5 i.xf2+
17 Bxf2 exd5 18 «xd5 ±a4 19 Bdfl ±c2 was
better for Black in Fressinet-Korchnoi, Cannes
1996) 13...f6 14 b5 £>a5 15 exf6 Bxf6 16 £>e5
JLe8 with approximately equal play, Bosch-
Lputian, Wijk aan Zee 1999.
8...£ff5 (D)
W
wm mm w% a mm
'<m&
Now we discuss:

18
French: Advance and Other Lines
Cl: 9i.e3 18
C2: 9±b2 21
Cl)
9 i.e3 £>xe3
Some other moves are also important:
a) 9...i.e7 10 i.d3 £>xe3 11 fxe3 f5 12 exf6
(White's dynamic options would be
considerably restrained if he leaves the centre closed)
12...±xf6 13 £>c3 (not 13 £>g5? ±xd4!: 14
#h5+ g6 15 ±xg6+ hxg6 16 #xg6+ &d7 17
#xe6+ &c7 18 #f7+ &b8 -+ or 14 exd4
#xd4 15#h5+&e7 16#f7+&d6+) 13...0-0
14 0-0 Wd8 15 #c2 h6 16 ladl ±d7 17 Wd2,
intending Jtbl/c2 and Wd3, was better for
White in Jonkman-Van Harreveld, Amsterdam
2000.
b) 9...±d7 10±d3(10£>c3?!£>xe3 11fxe3
£>xb4! 12 axb4 Jtxb4 leaves White in an
awkward pin; e.g., 13 Icl Ic8 14 #b3 #a5 15
&d2 0-0 16 ±d3 Ic7 17 £>gl f6!) 10...£>xe3
11 fxe3 ±e7 (1 l...f5!?) 12 0-0 Ic8 13 Wd2 0-0
14 £>c3 (D), and now:
bl) 14...£>xb4? 15 axb4 ±xb4 16 Wf2 Ixc3
17±xh7+!+-.
b2) 14...f6 15 £>a4 Wd8 16 exf6 Axf6 17
£>c5 b6 18 £>xd7 Wxdl 19 lacl ± Djurhuus-
Heim, Bergen 2001.
b3) 14...£>b8 15 *hl Wd8 (15...f5!? here
or on the last move is maybe better) 16 e4 Jtc6
17 b5 dxe4 18 £>xe4 ± Movsesian-M.Socko,
Koszalin 1998.
c) 9...f6!?(T>)andnow:
cl) 10 exf6 gxf6 11 ±d3 £>xe3 12 fxe3 ±h6
13 We2 i.d7 14 £>c3 £>e7 = Jonkman-San
Segundo, Mondariz Z 2000.
c2) 10 b5 can lead to a draw by repetition,
but there are attempts to avoid it, for Black at
least. 10...£ixe5! (10...£>a5 might not be bad
either) 11 dxe5 £>xe3 12 fxe3 #xe3+ 13 We2
(White can't really avoid the draw since 13
i.e2 leads to difficulties: 13...ix5 14 If 1 fxe5
15 Wd2 e4 16 #xe3 ±xe3 17 £>fd2 e5 18 £>c3
±e6 + Acs-S.Williams, Budapest 1994) and
now:
c21) 13...Wcl+ 14Wdl and then 14...ffe3+
has been the peace-settling conclusion of many
games, either here or after a few more checks.
14...Wb2 is one attempt to play for more, but 15
£>bd2 fxe5 16 Ibl Wc3 17 Ib3 Wc7 18 b6!?
axb6 19 i.b5+ i.d7 20 i.xd7+ &xd7 21 0-0
was better for White in Kunte-Cherniaev, Biel
2001.
c22) I would prefer Watson's idea of
exchanging queens with 13...Wxe2+ 14 Jtxe2
fxe5 15 £>xe5 ±d6 16 £>f3 0-0 17 0-0 ±d7 18
^bd2 Iac8. This is probably not so bad for
Black even though he only has two pawns for
the piece: he has good control of the centre and
his bishops are strong. This could maybe even
be improved upon by playing 13...Wcl+ 14
Wd\ and only then exchanging queens.
c3) 10 ±d3!? £ixe3 11 fxe3 fxe5 12 b5
£>xd4!? (12...e4 13 bxc6 i.d6 14 0-0 0-0 15
£>g5! h6 16 Ixf8+ ±xf8 17 £>xe4 dxe4 18
Jtxe4 bxc6 19 £M2 ± L.B.Hansen-Antonsen,
Farum 1993) 13 exd4 e4 14 ±xe4! dxe4 15
^e5 (D) and now:

ADVANCE: 5...Wb6 6 a3
19
c31) 15...g6?! 16 0-0±g7 17*hlixe5 18
dxe5 #xb5 19 &d2! #d3 (19...Wxe5 20 £lc4
Wd5 21 We2 looks very dangerous for Black)
20 Wcl! Bf8 21 Idl! (21 Ixf8+ 4>xf8 22
ffc5+ *g8 23 If 1 Ad7 24 Wf2 e3 25 Wf7+
4>h8 26 #f6+ =) 21...ffe3 22 lei Wf4 23
£\xe4 Wxcl 24 Saxcl ± Malvasio-Valli, e-mail
1999.
c32) 15...Wa5+!? 16 #d2 #xb5 17 £ic3
Wa6oo.
c33) 15...i.d7! 16 0-0 (16 #h5+ g6 17
£>xg6 hxg6 18 #xh8 0-0-0 looks like a good
exchange sacrifice) 16...0-0-0 17 £>c3 ix5 18
^a4 &xd4+ 19 *hl #d6 20 £rf7 #d5 21
£\xh8 Sxh8 with excellent compensation, Pot-
kin-Filippov, Russian Cht (Togliatti) 2003.
10 fxe3 (D)
10...g6
The plan of fianchettoing the dark-squared
bishop is sensible. Black has the long-term
advantage of the bishop-pair but suffers from
less space. However, he should be careful not
to attempt to free his position prematurely.
Instead, this keeps his position solid and prepares
to develop the bishop on h6, from where it will
tie White to the defence of the e3-pawn.
Alternatively: 10...i.e7 11 i.d3 - 9..±e7 10 &d3
£hxe3 llfxe3\ 10...Ad7 11 Ad3 - 9..Ad7 10
kd3thxe3 Ufxe3.
11 ±d3 (D)
11 £>c3 ±h6 12 Wd2 0-0 13 Icl f6 14 exf6
Ixf6 15 ^a4 #d8 16 ±b5 ^e7 17 0-0 £>f5 18
Icel £id6 19 i.d3 i.d7 20 £ic5 b6 21 £ixd7
Wxdl 22 ^e5 Ixf 1+ 23 Ixfl #a4 with good
counterplay, Dvoirys-Pliasunov, St Petersburg
2001.
lL..£.h6
A less-tested idea is just to play 1 l..JLg7!?
with the idea of a later ...f6 break, and this may
well be Black's best choice.
12#e2(T>)
This seems like the most logical as the e3-
pawn needs protection, but Morozevich has
played 12 ltt2!?. This has the advantage of
the knight being protected when it is
developed to c3, but a more far-sighted idea is that
the queen is more aggressively placed in
connection with a £\c3-dl-f2-g4 plan, which is
investigated more closely in the next note.
a) Morozevich-Milos, New Delhi FIDE KO
2000 went 12...i.d7 13 £>c3 £>e7 (13...0-0
should probably be met by a quick £\a4-c5 but
certainly not 14 0-0?? £\xd4!, as Kosten points
out) 14 g4! i.g7 15 0-0 Ic8 16 a4!? 0-0 (not

20
FRENCH: ADVANCE AND OTHER LINES
^...Wxl^? 17 £>b5 Wxd2 18 £>d6+ <£>d8 19
£ixd2 +-) 17 £ft>5! and White was better. Black
needs an improvement here.
b) Possibly 12...0-0 with a quick ...f6 is
Black's best reaction. It is useful to compare
this with the next note but one point is that on
d2 White's queen is blocking the way for the
knight in the aggressive line involving an early
g4.
EH 6
12...±d7
12...0-0 is a natural alternative, but castling
could prove premature. We then have:
a) 13 0-0 and now:
al) 13...a5 14 b5 £>e7 15 £>c3 (15 a4 avoids
Black fixing the queenside but allows 15...^f5
16 JLxf5 exf5 17 £>c3 JLe6 with a fine position
for Black) 15...a4 16 fcdl (16 Sabl ±d7 17
Sb4 is also very logical; I don't see a clear
refutation but 17...1rc7 18 £ixa4 Sa5!? is maybe
counterplay) 16...i.d7 17 £>f2 Sfc8 18 £>g4
Ag7 19 £>f6+ Axf5 20 exf6 £tf5 21 £>e5.
White's whole attacking concept looks
dangerous but referring to my comment to White's 12th
move, it is clear that White's queen would have
been much better on d2. From here the game
Kozel-Malakhatko, Yalta 1996 went 21...*d8
22 Wf2 Sc3 23 g4 £>d6 24 «h4?! Sxd3! 25
£>xd3 JLxb5 and Black had good
compensation, but I have a feeling that White doesn't
need to dig very deep to find an improvement
on this. One idea is 22 Wei!?, not allowing
Black's rook in on c3.
a2) 13...f6!? (to me this seems like a better
way of creating counterplay, one idea being 14
b5? £>xe5! 15 dxe5 fxe5 with very good play)
14 exf6 Sxf6 15 ^hl (this doesn't look
particularly necessary; better is the immediate 15
£>c3) 15...i.d7 16 £>c3 a6 17 £>a4 »c7 18 £>c5
Saf8, and Black has at least equalized, Berg
Jensen-Potkin, Khalkidhiki U-18 Ech 2000.
b) 13 g4!? Wd8 14 b5! (this is a more
accurate move-order than 14 h4 f6 15 exf6 Wxf6 16
£>bd2 Ag7 17 b5, when Black has 17...e5 18
bxc6 e4 with a messy position) 14...£>a5 15
£>bd2 f6 (Black reacts before his bishop
becomes in danger of being buried on h8 after a
white h4, g5 and h5-h6) 16 exf6 Wxf6 17 h4
±g7 18 h5 ± Aldrovandi-Castaldo, Italian Ch
(Saint Vincent) 1999.
13 £>c3£>e7 14 0-0 (D)
14...Sc8!
A clever preliminary move. On 14...^f5,
White can try 15 g4!?, when the position
becomes messy. Thipsay-Neelotpal, Indian Ch
(Mumbai) 2000 continued 15...i.xe3+ 16 *hl
±xd4 17 ^b5 i.xal 18 gxf5 exf5?! (18...0-0
19 Sxal ±xb5 20 ±xb5 exf5 is more critical)
19 £>d6+ *f8 20 Sxal h6 21 Scl with strong
pressure.
15 Sacl 4M5 16 Axf5
16 £>a4 might be better but then 16...*d8!
is fine for Black. Note that Black should avoid
16...Axe3+ 17 Wxe3 £ixe3 18 £>xb6 axb6 19
Sxc8+ ±xc8 20 Scl, when White has
excellent play for the pawn.
16...gxf5
Black is doing very well. In G.Wall-Lputian,
Montecatini Terme 1999 Black soon had an

Advance: 5..Mb6 6 aS
21
overwhelming position: 17 Sc2?! Wa6! 18
Hfcl? Sxc3! 191^6 ±xe3+ 20&fl Sxc2 21
Wxb7 Sxcl+ 22 &e2 f4 -+.
C2)
9 ±b2 (D)
Then:
C21: 9...i.d7
C22: 9...i.e7
21
25
Or9...a5 10 b5 a4 11 g4£>fe7 (ll...£>h6 12
h3 ±) 12 £>c3 ^b8 13 i.d3 (White has plenty of
time to pick up the a-pawn, but 13 £>xa4 Wd8
14 £>c3 £>g6 15 h4 h5 16 g5 is also good)
13...«ki7 140-0 4ig6 15 HclAe7(15...«d8!?)
16 £>xa4 lfa5 17 £>c5 and White is better,
Sveshnikov-Bareev, Moscow PC A rpd 1995.
C21)
9...i.d7 (D)
This is meant to prevent a white set-up with
JLd3 but the drawback is that White can now
immediately drive Black's knight away from
f5.
10 g4
This has been accepted as the critical
continuation. Other moves:
a) 10 £>c3?! £>cxd4 11 £>xd4 Wxd4 12 Wxd4
£xd4 13 £ixd5 £>c2+ 14 <£>d2 (14 *dl Aa4!
—) 14...Sc8 15 Scl £>xa3! 16 £>c7+ <£>d8 17
£.xe6+ i.xe6 18 i.xa3 Sxcl 19 *xcl a5 +
Donev-Gaertner, Voralberg 1997.
W
& A Wm JL Way, a vmz a
6'£ ^ '//////A - %%%£ 4%
l
wwm
w& wz
b) 10 ±e2 Sc8 11 0-0 ±e7 12 *hl 0-0 13
£>c3 £>a5 (13...£>cxd4?! 14 £>xd5 exd5 15
£>xd4 ±) 14 £>a4 i.xa4 (14...Wc6!?) 15 Wxa4
£>c4 16 Wb3 £>xb2 17 Wxb2 Sc7 ? Kalezic-
Knezevic, Vrnjacka Banja 1999.
c) 10 Ad3 £>cxd4 11 £>xd4 £>xd4 12 0-0 a5
13 £>c3 gave White some play for the pawn in
Pyhala-Votruba, Helsinki 1989 but I don't trust
it.
d) 10h4!?(D)andhere:
If ■#■
mi A 111 AAA
n n 1 a
dl) 10...h5 is the safer of Black's options. In
Cherniaev-Sedina, Davos 2004, Black obtained
a comfortable draw after 11 Ae2 Sc8 12 1U2
k.€l 13 *fl a6 14 *gl £>a7 15 £>c3 a5 16
bxa5 *xa5 17 Ad3 £>c6 18 Ac2 £>a7 19 Ad3
£k;6 20Ac2£>a7 21 Ad3.
d2) 10...Sc8 11 h5!? (11 g4 £>fe7 12 £>c3
£>a5! 13 £>a4 »c6 14 £>c5 £>c4 15 Acl h5 16
gxh5 £tf5 17 JLh3, Movsesian-Haimovich,
Panormo ECC 2001, and now Movsesian gives

22
FRENCH: ADVANCE AND OTHER LINES
17...b6 as fine for Black) ll...&xb4!? (Il...£>a5
is the alternative) 12 axb4 *xb4+ 13 *d2 Sc2
14 Aa3 Wxd2+ 15 £>fxd2 £>xd4 16 Sh4 Sxd2
17 <A>xd2 £>b3+ 18 <4>c3 £>xal 19 <4>b2 f6 20
exf6 gxf6 21 Axf8 *xf8 22 <A>xal Sg8 23 h6
and White skilfully won the endgame in Nayer-
Kraai, Las Vegas 2004, although I am far from
sure how to assess the position at this moment.
We now return to 10 g4 (D):
10...£>fe7
A slightly provocative retreat: Black's knight
goes via h6 to f5, only to land on e7, before the
bishop is developed! After 10...£}h6, White has
two ways of defending his g-pawn:
a) 11 Sgl f6 (ll...Sc8!?, intending ...£>a5)
12 exf6 gxf6 13 £k3 £tf7 (D) and now:
al) 14&a4Wc7 15HclWf4 16&c5Axc5
17 dxc5 £ke5 (Psakhis suggests 17...e5 as a
possible improvement, but then 18 ^xd5 Axg4
19 Ac4 Sf8 20 Ae2! h5 21 b5 £kd8 22 c6
might leave Black in real trouble) 18 £ixe5 and
now 18...^xe5 19 Sg3! was complicated but
probably good for White in Short-Lputian,
Batumi Echt 1999. Instead 18...fxe5 19 Sg3
Sf8 might be a little annoying for White.
a2) 14 Scl Ah6 15 Sc2 £>e7 16 h4 Af4 17
Ad3 h6 18 Se2 *c7 19 *b3 £>d6 20 Sc2 *b8
21 £te2 £te4 22 b5 a6 23 bxa6 V2-V2 Lautier-
Bauer, French Ch (Val d'Isere) 2002.
b) 11 h3 (D) and now:
bl) ll...f6 12 £k3 (12 exf6 gxf6 13 £k3
thfl 14 £}a4 followed by £k5 t is given by
Lputian, although this was intended with the
move-order 12 Qsc3 £hf713 ex/69 etc.) 12...fxe5
13 dxe5 Ae7 14 £>a4 ttd8 15 Scl 0-0 16 £>c5
Axc5 17 Sxc5 £>f7 18 Ag2 b6 19 Scl £>e7 20
^d2 £}g6 was fine for Black in Sveshnikov-
Lputian, Sochi (3) 1993.
b2) ll...£>a5 12£>bd2(12 Acl!?) 12...£>c4
13 Axc4 dxc4 14 £>xc4 Wb5 15 £>fd2 Ac6 16
0-0 Ae7 with compensation for Black, Prasad-
Harikrishna, Hyderabad 2000.
b3) 1 l...Ae7 12 Ad3 Sc8 13 £>bd2 £>b8 14
lfe2 a6 15 £>bl! £k6 16 0-0 0-0 17 £k3!
£>xd4? (17...!tt8 {Timman} is more sustained)
18 £>xd4 *xd4 19 £>e4 *b6 20 £>f6+ Axf6
21 exf6 Ab5 22 Axb5 *xb5 23 !fe3 left Black
in severe difficulties in Timman-Cu.Hansen,
Malmo/Copenhagen (2) 2003.
b4) ll...Sc8 12 £>c3 £>a5 13 £>a4 *c6 14
Scl (14 £>c5 £k4 15 Ac 1 is a natural
alternative) 14...£k4 15 Bxc4?! (although interesting,
this idea probably doesn't stand up to close

ADVANCE: 5..Wb6 6 a3
23
scrutiny) 15...dxc4 16£>c3 Wb6 17 Ag2a5! 18
b5 Axb5 19 d5 Ac5 20 0-0 Bd8 21 d6 Ac6
gave White practical chances due to the strong
centre pawns and Black's misplaced knight on
h6 in Shirov-Sadvakasov, Astana 2001, but with
correct play it shouldn't be quite enough to
compensate for the material losses.
11 £>c3 (D)
ll...£&5
The most common move. Black
immediately tries to exploit White's weaknesses on the
queenside. Other moves:
a) ll...Wd8 12 Ad3 £>g6 13 h4 Ae7 14 g5
€M4 15 Ac2 Bc8 16 Wd2 £>h5 17 Ad3 g6 18
£te2 ± Savic-Damljanovic, Herceg Novi 2001.
b) 1 l...£>g6 12 £>a4 (12 h4 h5 13 g5 might
be more accurate) 12...»d8 13 h4 h5 14 g5 b6
15 Ad3 £>f4 16 Aa6 £>e7 17 £>c3 £>f5, Neved-
nichy-Lputian, Ohrid Ech 2001, and now 18
tfcl! looks strong, with the idea 18...£>g6 19
£>b5!. Black's best try is maybe 18...b5!? 19
*xf4 «T>6 but 20 Axb5 Axb5 21 £>xd5 should
be good for White.
c) ll...h5 12 g5 £>g6 13 £>a4 Wd8 14 Bel
Ae7 15 Bgl 0-0 16 h4 a5 17 b5 £>a7 18 Wb3
£>c8 19 Ac3 £>xh4!? 20 £>xh4 Axg5 21 Vdl
Axel 22 *xh5 Ae8 23 Ad3 f5 V2-V2 Peng
Xiaomin-Korchnoi, Calcutta 2000. White forces
a draw with 24 exf6 Axh5 25 Bxg7+ <±>h8 26
Sh7+, etc.
12&d2
Or:
a) 12&a4«fc613b5(13£>c5£>c414Acl
h5 =) 13...»c7 14 £>d2 £>c4 15 £>xc4 dxc4 16
£>c3 £>d5 17 Ag2 £>xc3 18 Axc3 Axb5 19 d5
exd5 20 »xd5 Ac6 (20...Aa6!?) 21 »xc4 Bc8
= Lindfeldt-P.H.Nielsen, Danish Ch (Nyborg)
2001.
b) 12 *c2!? £>c4 13 Axc4 dxc4 14 £>d2
£>d5 (14...»c6 15 £>ce4 £>c8 16 £>xc4 b5 17
£>e3 »xc2 18 £>xc2 Ac6 19 f3 ± S.Ivanov) 15
£>xc4 »c6 16 £>e4 £>b6 17 £>cd6+ Axd6 18
£>xd6+ <±>e7 19 *xc6 Axc6 20 Bgl £>c8 21 b5
(21 d5! is a much better chance; e.g., 21...Axd5
22 Bel £>xd6 23 Bc7+ <±>e8 24 exd6Bg8 25 g5
Bd8 26 Ae5 Bd7 27 Bg4! ±, with the idea Bh4)
21...Ad5 22 Bel £>xd6 23 Bc7+ <±>d8 24 exd6
Bc8 25 Bxc8+ <±>xc8 26 Ac3 <±>d7 27 Ab4 Bc8
28 <±>d2 V2-V2 Alekseev-S.Ivanov, Tomsk 2001.
There are no real chances for White to make
progress in view of the opposite-coloured
bishops.
12...Bc8 (D)
13 Bel
Occupying the c-file is logical, of course.
White's idea is that when Black's knight goes to
c4, White will capture with the bishop and
follow up with £>ce4. But the b2-bishop is still
undefended, and thus Black doesn't need to play
...£>c4atonce.
13 £>a4 *c6 14 b5 *c7 15 Bel *d8 has a
bad reputation in view of the game Relange-
Anic, Sabac 1998, which went 16 Ad3?! Bxcl
17 Wxcl £>g6 18 £tf37! £>c4 19 Axc4 dxc4 20
tfxc4 a6! with a large advantage to Black, but
this can easily be improved upon. One idea is
16 Ac3!? with the idea of Ab4. Another is 16
£>c5, which might be White's best. Here are

24
French: Advance and Other Lines
two sample lines, which are both difficult to
assess:
a) 16...£>g6 17 £>xd7 f xd7 18 h4 h5 19
2xc8+fxc8 20g5oo.
b) 16...h5 17 Ac3 (17 gxh5 £>f5 18 i.d3 •)
17...£>g6 18 £>xd7 <£>xd7 19 gxh5 £>f4 20 ff3
Wg5co.
13...£>g6! (D)
This highlights weaknesses on the kingside.
If Black could succeed in exchanging a few
pieces and complete his development he would
have excellent chances of a good game. But for
the moment he has to deal with White's large
space advantage. Two alternatives to the text-
move are worth mentioning:
a) 13...^c4 (this may be a viable option but
there seems to be no reason to play it this
early) 14 i.xc4 dxc4 15 £>ce4 £>d5 16 £>xc4
Sxc4 17 Sxc4 Ae7 18 0-0 0-0 19 Sc5!? and
White should be better, Jonkman-Vysochin,
Cairo 2002.
b) 13...h5!? 142c2£>c4(14...£>ac6 15^a4
fd8 16 £>c5 is better for White) 15 ±xc4 dxc4
16 £>ce4 £>d5 17 £>xc4 Sxc4 18 Sxc4 a5 19
0-0 axb4 20 axb4 lfa8 with unclear play, Ivan-
chuk-Bareev, Dubai 2002.
c) 14 i.b5 (D) and then:
W
HPJb^%2& W&.
, hi iHA»fliiiii
&ii fif 4 if Hi
PS PI PA,
Wf.""/'w^.'"/'mm. %•"•
Wb&%
AHB
14 h4
Kruppa and Komarov have queried this but
that may be unjustified, and I see none of the
alternatives as any better:
a) 14 i.al £>c4 15 i.xc4 dxc4 16 £>ce4
JLb5 looks fine for Black.
b) So does 14£>a4Hxcl 15 Wxcl (15±xcl
#xd4 +) 15...i.xa4 16 bxa5 Wc6.
y/A •
M2 _^
wd"&WB 4%f ''WB
"' '"" A/m Am
IB
cl) 14...i.xb5 15 bxa5 (15 £>xb5 £>c6 ?)
15...f xa5 16 ^b3 (16 £>xb5 Sxcl 17 *xcl
fxb5 18 fc8+ <£>e7 19 Ac3 f6! creates an
escape-route for Black's king to avoid a perpetual
check) 16...Wb6 17 £>xb5 Bxcl 18 f xcl fxb5
19 fc8+ <£>e7 20 a4!? fxb3 21 fc7+ =.
c2) 14...£ic6 15 £>f3 (15 £>b3 Ae7 16 £>c5
JLxc5 17 bxc5 ^a5 appears fine for Black)
15...i.e7!? (not 15...£>cxe5? 16 dxe5 i.xb5 17
£>xb5 Bxcl 18 f xcl f xb5 19 fc8+ <£>e7 20
±d4 b6 21 ±e3 and White is much better) 16
h4?! (16 Ae2!? 0-0 17 £>a4fd8 18 £>c5 ±xc5
19 Bxc5 f6 also leads to favourable
complications for Black) 16...£>cxe5! 17 dxe5 i.xb5 18
h5 £>f8 19£>xb5 Bxcl 20*xcl fxb5 21 *c8+
i.d8 22 <£>d2 £>d7! + Alekseev-S.Ivanov, St
Petersburg 2001.
14...i.e7 (D)
Psakhis also suggests the unclear sacrifice
14...f xd4!? 15 bxa5 £>xe5 16 ±e2 f f4.
15 h5
Grishchuk criticizes this, perhaps rightly,
and it does indeed give Black extra room on
the kingside in the form of the g5-square.
Instead, the game Grishchuk-Radjabov, Wijk
aan Zee 2003 saw the stronger 15 g5! h6 16
gxh6 Bxh6 (16...gxh6!?) 17 h5 £>h4 18 fg4
&f5 19 ±d3 with a clear advantage to White,
as 19...£>xd4 20 f xg7 Af8 21 Wg8 would be
too dangerous.
15...£>f4
15...£>h4!?.
16fT3

ADVANCE: 5..Mb6 6 aS
25
w
■fm wm m. mm m
jl*..
A
■ AB
^
^f ii
1 fi
16£>a4? Sxcl 17 Wxcl (on 17 i.xcl, Black
has 17...Wxd4 18 i.b2 £>d3+ 19 i.xd3 Wxd3
20 bxa5 Wb5!, regaining the piece) 17..Jbca4
18 bxa5 Wxa5 19 Wc8+ i.d8 20 Wxb7 0-0 +
Psakhis.
16...±g5 (D)
w
i
p m
i
AB HA
murm
i pi I
"Wo"
We are following Grishchuk-Kruppa, Elista
2000 which now continued 17 £)e2 Sxcl+ 18
■£)xcl 0-0, and Black was better. From the
diagram I am not entirely convinced that White
should be in trouble. A few ideas:
a) 17 Sc2 and then:
al) 17...£ic6 18&a4ttd8 19£>c5 (19#dl
0-0 20 £>c5 Sc7 21 £>f3 f6 22 exf6 i.xf6 •)
19...b6!? 20 £>b7 (not 20 £ixd7? &xd4! -+)
20...£ixd4! 21 Sxc8 £>xf3+ 22 £>xf3 i.xc8 23
-iixd8 i.xd8 24 i.cl g5 25 hxg6 £>xg6 26 i.h6
X.
a2) 17...£>c4 18 £>xc4 dxc4 (Black could
consider 18...Sxc4!?) 19 £>e4 ix6 20 h6 gxh6
21 Sxc4, again with a position which is
difficult to assess.
b) 17h6!?g618 2c2^c419£>xc4dxc4 20
£>e4 ix6 21 Sxc4 0-0 22 Sc5 (22 Sxc6? Wxc6
23 £>xg5 Wxf3 24 £>xf3 Sc2 -+) 22...«to8 •.
C22)
9...i.e7 (D)
This, apart from developing of course, helps
the f5-knight, as g4 can now be met by ...£ih4.
W
10 h4
This renews the idea of playing g4, and
gains space before completing development.
The other main option is 10 JLd3:
a) 10...i.d7 11 0-0 and now:
al) 1 l...£>fxd4 12 £>xd4 £>xd4 13 Wg4 £>b3
14Wxg7(14Sa2!?) 14...0-0-0 15Sa2Sdf8 16
Wg4±
a2) 11...0-0 12 £>c3! £>cxd4 13 £>xd4 Wxd4
(13...£>xd4 14£>xd5 exd5 15 ±xd4±) 14 Axf5
Wxdl 15 Sfxdl exf5 16 £>xd5 ±.
a3) ll...g5 12 Axf5 exf5 13 £>c3 i.e6 14
£>a4 Wb5 (14...Wd8 15 Scl ±) 15 £>c5 g4 16
£>el JLxc5 17 dxc5 ± Fedorowicz-Benjamin,
Wijk aan Zee 1989.
b) 10...a5!?(Dj and here:
bl) 11 b5!? £>cxd4 (ll...a4!?) 12 £>xd4
£>xd4 13^c3 (13 Wg4 £>xb5 14 Wxg7 Sf8 •)
13...a4 14 0-0 ± H.Akopyan-Bhat, San
Francisco 2000.
b2) HWa4!?0-0 12b5h6 13 0-0(13i.c3!
£>b8 14 £>bd2 appears more to the point)
13...£>h4 14 £>xh4 i.xh4 15 Sa2 £>a7 16 £>c3

26
FRENCH: ADVANCE AND OTHER LINES
f5 * Aguilera-de la Villa, Spanish Cht (Pon-
ferrada) 1992.
b3) 11 Axf5 exf5 12 £>c3 Ae6 13 b5 a4! 14
bxc6 (on 14 0-0 both 14...£>a5 and 14.. .£rt>8 are
alright for Black, and the shortcomings of 14
32b 1 £ia5 15 #xa4 were illustrated in the game
Garrido Fernandez-Moskalenko, Barbera del
Valles 2003, where White quickly went down:
15...0-0 161U1 £>c4 17a4£>xb2 18Hxb2Wa5
19 #b3 Hfc8 20 lc2 lxc3 21 lxc3 Bc8 0-1)
14...1rxb2 15 £>xa4!? (15 0-0 bxc6!? 16 £>xa4
#b5 17 £>c3 #c4 was fine for Black in Shirov-
Khalifman, Linares 2000) 15...2xa4 16 cxb7
#c3+! 17£>d20-0 18 Ibl i.b4! (18...1xd4 19
b8# lxb8 20 lxb8+ if8 21 Hb3 is not totally
clear but probably good for White) 19 Uxb4 (19
axb4 32a2 gives at least sufficient counterplay)
19...1xb4 20 axb4 #xd4 21 0-0 Bb8 22 #a4
#b6 23 £>f3 #xb7 24 Ibl ±/=
We now return to the position after 10 h4
10...i.d7
Now White will gain even more space, which
is slightly awkward but Black hopes that White's
expansion on both sides will give him a chance
to counterattack. Other options for Black:
a) 10...a5!? 11 b5 a4 12 g4 £>h6 (D) and
al) 13&c3?!&xg414&xa4*a5+15£id2
#xa4 (15...^cxe5! might be a surprising
tactical resource, with the idea 16 £>c3 Wall
followed by ...h5 and White doesn't have time to
capture either of the knights) 16#xg4^xe5 17
#xg7 £>g6 18 h5 Af8 19 Wf6 Ae7 20 Wgl
Af8 21 Wf6 Ae7 22 #g7 V2-V2 Sandipan-Dol-
matov, Moscow 2002.
a2) 13 Sgl!? ^a5 with two options for
White:
a21) 14 £>c3 £>b3 15 Ba2 #a5 16 Ad3 Ad7
17 £M2?! (probably based on a
miscalculation; something like 17 itfl was called for)
17...£>xd4 18 £>xd5 Axb5! 19 Ac3 WdS 20
£>xe7 Axd3 21 Ab4 £>g8! (fantastic; the d5-
square is what Black needs) 22 £>xg8 #d5 23
Sg3 #hl+ 0-1 Emelin-Dolmatov, Russian Ch
(Krasnoiarsk) 2003. A picturesque final
position.
a22) 14 #xa4!? Ad7 15 £>c3 0-0 and now
16 Vdl 2fc8 17 £>a4 #d8 18 £>c5 £>c4! wasn't
bad for Black in Lavrov-Sambuev, Tomsk 2003,
but 16 #c2! looks more to the point, intending
^g5.
b) 10...h5 11 Ad3 (D) and then:
bl) ll...a5 and now:
bl 1) 12 Axf5 exf5 13 £>c3 and here:

ADVANCE: 5..Mb6 6 a3
27
bill) 13...axb4?! 14 £fcd5 #a5 15 £ke7
b3+ 16 £>d2 £>xe7 17 #xb3 ± Sveshnikov-
Bareev, Russian Ch (Elista) 1996.
bl 12) 13...i.e6 (this is similar to 10 &d3 a5
11 Jbc/5 exf5 12 £hc3 k.e6 but with the
inclusion of h4 and ...h5, which should in general
favour White). There are some interesting recent
developments in this line: 14 b5 a4 15 #d3
£>a7 16 0-0 2e8 17 Acl Sc4 and after 18 Idl
£>xb5 19 £>e2 #c6 20 Ag5 Axg5 21 £>xg5
£k3! Black won a pawn in Grishchuk-Lputian,
Bled OL 2002, but the very next day White
improved with 18 £>e2! #xb5 (18...£>xb5 can still
be tried; then 19 Ibl #c6 20 £>d2 f4 21 £>xc4
dxc4 must be analysed) 19 Ag5 #b3 20 #xb3
axb3 21 i.xe7 4>xe7 22 Ifbl in Vysochin-
Polivanov, St Petersburg 2002: White wins back
the pawn with the better endgame in prospect.
bl2) 12 #a4! (this recent finding of Sandi-
pan's might cast doubt on Black's play in this
line) 12...Ad7 13 b5 £>a7 14 £>c3 #d8 15 ttdl
g6 16 g3 a4 17 Acl £>c8 18 Ag5 £>b6 19 0-0 ±
Sandipan-Barua, Indian Ch 2004.
b2) Il...g6 12 Axf5 gxf5 13 £>c3 Bg8 14
0-0! (14 £>g5 #xd4 15 #xd4 £>xd4 16 £>xd5
£ic2+ 17 <A>d2 exd5 18 *xc2 Axg5 19 hxg5
Ixg5 20 Ad4 Ae6 21 g3 *d7 22 Ae3 Bg4 23
Af4 lgg8 24 4>d3 lgc8 25 lacl Sc4 gave
Black enough counterplay in Grishchuk-Sakaev,
Tomsk 2001) 14...Ig4 15 £>e2 Ad7 16 g3 ±
Grishchuk-Zhang Pengxiang, Shanghai 2001.
11 g4
This is logical, but Bangiev gives 11 h5! (his
mark). I am not sure about this but it is surely an
interesting idea to wait and see what Black
does. Il...lc8(ll...f6?! 12i.d3!fxe5 13 Axf5
exf5 14 dxe5 ±) 12 #d2 0-0 13 Ad3, and now
13...£>cxd4 14 £>xd4 £>xd4 15 Axd4 #xd4 16
Axh7+ *xh7 17 #xd4 lcl+ 18 *e2 Ixhl led
to an unclear position in Motylev-Miljanic,
Bucharest 2000. Also interesting is Bangiev's
13...a6, and then his main line runs 14 Ac2! f6!
15 g4 (I think 15 Axf5 exf5 16 £k3 Ae6 17
£te2 should be seriously considered) 15...£Mi6
16 exf6 (16 #d3?! f5) 16...1xf6! 17 #d3 lxf3!
18 #xf3 £>xd4 19 Axh7+ *h8 W, but Black
might in fact be better in the final position.
Il...£>h6 12 Sgl (D)
12...0-0!?
Another interesting idea is 12...f6 13 exf6
gxf6, and now Svidler-Beliavsky, Novosibirsk
1995 went 14 Ad3 £>f7 15 #e2 0-0-0 16 £>c3
<A>b8 17 lei lc8 18 £>a4 #d8 19 £k5 Axc5
20 lxc5 £>d6 21 g5! <4>a8 22 b5 £>a5 23 gxf6
#xf6 24 £>e5 Wei V2-V2, but it is not clear that
this is all that comfortable for Black. In Zenker-
Lannaioli, e-mail 2000, White was successful
with a quicker £k3-a4-c5 which I believe
should promise White the better chances: 14
£k3 Bc8 15 £>a4 #c7 16 £>c5 £>f7 17 g5.
13 £k3 f6
Obviously, Black cannot afford to play too
reluctantly. This break involves a sort of
thematic exchange sacrifice. In return Black
obtains better space for his minor pieces. After 14
£}a4, Grishchuk-Johannessen, Reykjavik 2000
continued 14...#d8 15 exf6 lxf6 16 g5 lxf3
17 #xf3 £tf5 18 lg4, and Black didn't seem to
have enough for the exchange. 14...#c7 is a

28
FRENCH: ADVANCE AND OTHER LINES
possible improvement, which should be
investigated. If White continues in the same way,
various ...£>xd4 ideas are more appealing than
with the queen on d8, and the back rank is free
for Black's rook, which may join the game
from f8.
D)
6...c4 (D)
mXwwmwM
YAW/A ~ V/////A V/////A m. V/yi^/A m.
This advance highlights an obvious
drawback of White's last move, as b3 was
weakened. The situation has now changed as the
centre is closed, and the board is now divided
into two obvious battlefields, with White
usually trying to exploit his space advantage on the
kingside and Black in turn aiming to invade
White's queenside on the light squares.
However, it is not seldom to see play develop along
different lines, with, for example, White trying
to open the queenside, or Black aiming to seize
the initiative on the kingside, with a counter-
strike when White advances.
7&bd2
This is White's most flexible. White will
normally need to choose between a fianchetto of his
light-squared bishop or simply developing to e2,
but £>bd2 almost always fits in with White's
plans. Meanwhile, White might also think about
a quick opening of the queenside with b3 if he is
allowed. Moreover, 7 £>bd2 sets a little trap as
the typical French move 7...£>ge7(??) can be
met by 8 JLxc4!, netting a pawn because 8...dxc4
9 £\xc4 is decisive. Quite a few games have seen
this tactic, and about as many have also seen
White miss it. The most noteworthy 4misser"
was Kasparov in a simultaneous game, and
Yusupov is the most famous victim of the idea,
in Illescas-Yusupov, Linares 1992, albeit with
the moves .. JLd7 and g3 inserted.
A brief look at White's alternatives:
a) 7 JLe2 will most likely transpose to other
lines; e.g., l..±dl 8 0-0 £>ge7 9 £>bd2 £>a5 - 7
*hbd2 tha5 8 ±e2 k.d7 9 0-0 &e7, or 7...f6 8
0-0 fxe5 9 £>xe5 £>xe5 10 dxe5 Ac5 11 £>d2 -
7 *bbd2 j6 8 ke2 fxe5 9 *hxe5 Z&xe5 10 dxe5
&c5 11 0-0.
b) 7 h4 Ad7 8 h5 (8 g3 f6! is similar to 'c'
below) 8...f6! 9 !fe2 £>a5 (9...fxe5 10 £>xe5 £>xe5
11 *xe5 £>h6!? is also fine for Black) 10 Af4
£>b3 11 Ba2 0-0-0 12 g4 ^h6 13 ±h3 £tf7 14
0-0 g5! 15 Ae3 Ae7 16 exf6 Axf6 17 £>e5 Bhf8
18 £>xd7 Bxd7 19 £>d2 £>xd2 20 *xd2 Vd6 21
Baal JLd8 followed by ...JLc7 was better for
Black in Djuric-Vaganian, Bled/Roga§ka Slatina
1991.
c) 7 g3 f6!? (this seems like a good time to
make a break against the centre as White is
committed to fianchettoing; 7...£>a5 8 £>bd2 -
7 $Sbd2 foa5 8 g3) 8 exf6 (8 Ah3 fxe5 9 £>xe5
£>xe5 10 dxe5 Ac5 11 #h5+ g6 12 !fe2 ±d7
13 £>d2 £>e7 14 £>f3 h6 15 0-0 0-0-0 16 Bel
£>c6 17 Wc2 Bdf8 18 Ag2 g5 was fine for
Black in Malaniuk-Uhlmann, Tallinn 1987)
8...£>xf6 9 JLg2 (a quick attack against e6
yields nothing as 9 Ah3 Ad6 10 lfe2 0-0 11
i.xe6+ &h8 12 i.xc8 Baxc8 13 0-0 Bce8 gives
Black plenty of compensation) 9...JLd6 10 0-0
0-0 11 !fe2 (11 £>bd2?! allows Black to free
himself with ll...e5!) Il...*h8 12&e5&d7 13
£>xc6 (13 f4 £>dxe5 14 fxe5 Bxfl+ 15 <&xfl
{otherwise ...£>xe5} 15..JLe7 = Uhlmann)
13...Wxc6 14 f4 £>f6 15 £>d2 ±d7 16 £>f3
£>e4 = Gertler-Zsu.Polgar, New York 1985.
We return to 7 £>bd2 (D):
Now:
Dl: 7...f6 29
D2: 7...£>a5 31
As mentioned above, the natural 7...£>ge7??
is a common blunder in view of 8 JLxc4!.
7.. JLd7 has for some time been considered
inaccurate in view of 8 b3!? cxb3 9 £>xb3 (9

ADVANCE: 5..Mb6 6 a3
29
i.b2 £>a5 - 7...*ha5 8 b4 cxb3 9 k.b2 k.d7)
9...£>a5 (9...Sc8 10 Ad2 ±) 10 £>xa5 lfxa5 11
JLd2 (D) and now Black has chosen between:
a) ll...#a4 12 *bl! ix6 13 ±d3 £>e7 14
0-0 h6 15 Scl £ic8 16 c4 ± Sveshnikov-Tim-
man, Tilburg 1992.
b) ll...Aa4 12 Wbl (the alternative 12 c4
Axdl 13 ±xa5 ±xf3 14 gxf3 could also be
investigated) 12...Wc7 (12...Wb6!?) 13 Ad3 £>e7
14 0-0 Wd7 15 Scl (White adopts Sveshnikov's
plan from above, but Kosten also mentions the
idea of playing 15 £>g5 h6 16 £>h3, intending
£rf4-h5, or f4-f5, but to be fair Black could
possibly prevent this by playing ...h6 before he
develops his knight) 15...h6 16 Sa2 £k8 17 c4
dxc4 18 Sxc4 Ac6 19 Sxc6!? (else White
would be struggling positionally) 19...bxc6 20
d5! *xd5 (20...exd5 21 e6! *xe6 22«b7 £>b6
23 Ag5 hxg5 24 Se2 lfxe2 25 lfxc6+ <&d8 26
Axe2 ±) 21 Ae4 *d7 22 Sc2! £>e7 23 Sb2,
A.Praznik-Yakimenko, corr. 1993-6, and now
Praznik mentions that Black should invite a
repetition with 23...Sc8 24 Sb7 Sc7 25 2b8+
Sc8 26 Sb7, etc.
Obviously, White doesn't have to engage in
the line with 8 b3!? but fianchettoing the light-
squared bishop is still a problem, as 8 g3 f6! is
similar to 7 g3f6!. An alternative is 8 Ae2 £>a5
- 7 GSbd2 £ha5 8 &e2 &d7 (Line D2).
Dl)
7...f6 (D)
This direct break is probably better suited
against the immediate fianchetto but is still
fully playable, and in recent years Pelletier has
taken it up with success.
8±e2
The theoretical recommendation.
Alternatives are seen more rarely:
a) 8 exf6 £M6 9 g3 is similar to 7 g3f6 8
exf6 tfox/6 but here White's knight is
prematurely placed on d2.
b) 8 g3 fxe5 9 £>xe5 £>xe5 10 dxe5 ix5 11
«h5+ g6 12 !fe2 £>e7 13 Ag2 (13 Ah3 Ad7 -
7 g3f5 8 ±h3fice5 9 Z&xe5 foxe5 10 dxe5 &c5
11 ««+ g6 12 y&e2 Ad7 13 fod2 Zhe7)
13...i.d7 14 0-0 0-0-0 15 £tf3 h6 16 Ae3!? and
after 16...Axe3 17 *xe3 *xe3 18 fxe3 White
might claim a tiny edge in view of Black's bad
bishop but with accurate defence Black should
hold, Honfi-Uhlmann, Solingen 1974.
However, this is not much fun for Black to play, and
I would still prefer Brynell's 16...£te6, after

30
FRENCH: ADVANCE AND OTHER LINES
which 17 ±xc5 «xc5 18 Sadl g5 19 £>d4
£\xd4 20 Sxd4 Sdf8 was equal in J.Johans-
son-Brynell, Swedish Ch (Lindesberg) 1993.
c) 8 b3 fxe5 9 bxc4 e4 10 £>g5 i.e7 11 h4
^f6 ? Mary ska-J.Urban, corr. 1984.
d) 8 b4 fxe5 9 £>xe5 £rf6 =.
e) Little attention has been paid to the
aggressive move 8 £\h4!?, threatening Wh5+, or
simply wanting to follow up with f4. Dvoirys-
Vallin, Cappelle la Grande 2001 continued
8...g6 9 f4 fxe5 (9...±d7, intending ...0-0-0 is
maybe a better idea) 10 fxe5 ±g7 11 £\hf3
£>h6? (Il...£>a5 is indispensable) 12 ±xc4!
dxc4 13 £>xc4 Wa6 14 £>d6+ <&d7? 15 ±xh6
±xh6 16 d5!? exd5 17 Wxd5 with a winning
attack.
8...fxe5 (D)
W
9£>xe5
9 dxe5 might also be playable but White
usually takes with the knight in such positions to
let himself have a little more space.
9...£>xe5
9...£tf6 is worse; 10 f4 gave White a strong
bind in Pahtz-Uhlmann, Nordhausen 1986.
10 dxe5 ±c5
10...£>h6? 11 £>xc4! dxc4 12 ±xh6 netted a
pawn for White in Pahtz-Uhlmann, Erfurt 1985,
as 12...gxh6? 13 ±h5+ <£e7 14 Wf3 <£d7 15
0-0-0+ <£c7 16 Wf6 is decisive.
110-0£>e7(D)
Again 11...£\h6?! is not advisable: 12£\xc4!
dxc4 13 ±xh6 0-0 (13...gxh6? 14 ±h5+ <&e7
15 Wg4 +-) 14 Ae3 ±xe3 15 fxe3 «xe3+ 16
<£hl±.
W
« *«
'""''"«"'■ A
kA&w
1 13-fi.il
*!
12 b4
This has been considered the critical move
but 12 ±h5+!? g6 13 ±g4 is another idea. By
weakening the dark squares on the kingside,
White hopes to be able to take control of them
after £tf3. Moreover, it may be risky for Black
to castle queenside as White can open this front
with b3. 13...h5!? (this looks risky) 14 k&2
i.d7 15 Sbl 0-0!? (15...0-0-0 16 b4 cxb3 17
£\xb3 might be good for White) 16 £\xc4 dxc4
(16...Axf2+ 17 *hl *c7 18 Ag5 ±) 17 Wxd7
Sad8 18 Wa4 Sxf2 19 Sxf2 Axf2+ 20 *hl
Wc6 21 Wxc6 £>xc6 22 ±g5 Se8 23 Af6 was
better for White in Vallejo Pons-Pelletier, Biel
2002.
12...cxb3 13 £>xb3 0-0 14 £>xc5 Wxc5 15
Wd3!? (D)
This is slightly more aggressive than the
alternative, 15 Wc2, since White can now in
some lines use his queen actively on the king-
side.
. m m
***** m 11
zmfk Vwmy, **
i* i
IBA!
mail i
AB ■
m
mm « «
^f# ^^^ ^^^ ''*&'$&&

ADVANCE: 5..Mb6 6 a3
31
15..J?c7
Black's dream would be a transfer of his
light-squared bishop to g6 but this is not easy.
Pelletier gives 15...±d7?! 16 a4 ±e8 17 ±a3
Wc7 18 Wh3, which is clearly better for White.
Instead he prefers 15...If7: 16 ±e3 Wc7 17 f4
Ml =, and if 16 a4, 16...#c7 is similar to the
text-move but with White having spent time on
a4 which is maybe not that useful here.
16 ±g5!
A strong improvement on 16 f4?! Jtd7 17
Ae3 ±e8 18 ±g4 Wc6 followed by ...±g6,
when Black was better in Ilijin-Huss, Biel open
2000.
16...If 7 17 f4
Now:
a) After 17...±d7 (Grishchuk-Pelletier, Biel
2001) Pelletier thinks White is better following
the prophylactic 18 ^hl, which usefully
removes the king from the a7-gl diagonal.
b) Pelletier gives 17...£tf5 as better, with the
idea 18 g4 (18 Ah5!?) 18...Wc5+, when after
19 *hl £>e3 20 Hf3 £>xg4 21 Wd4 Wxd4 22
cxd4 Ic7 Black is doing OK. Instead, 19 If2
looks critical; e.g., 19...h6 20 gxf5 hxg5 21 f6
with a strong initiative for White, or 19...£te3
20 f5.
D2)
7...£>a5 (D)
White has two main options:
D21: 8±e2 31
D22: 8g3 33
Despite Black's efforts to prevent it, 8 b4 is
seen once in a while. After the forced 8...cxb3,
White would like to open the position further
with c4, but must first prevent Black from
playing 9...b2:
a) 9 ±b2 ±d7 10 c4 dxc4 11 £>xc4 £>xc4
12 ±xc4 £>e7 (12...±b5 13 £>d2 ±xc4 14
£\xc4 Wc6 may also be good) 13 0-0 ±c6 is
equal.
b) 9 Ibl ±d7 10 c4 dxc4 11 £ixc4 £>xc4 12
±xc4 Ic8 13 Wd3 Wc7 (13...±xa3 14 Ixb3
±b4+ 15 ±d2 ±xd2+ 16 £>xd2 «) 14 ±xb3 a6
? Gaudron-Gervasio, Le Touquet 1998.
D21)
8±e2
The fianchetto of this bishop has become
more popular, but the text-move has its pluses
too. One of the ideas is that in some lines
White can seek to prevent Black's attempt at
blockading the queenside by playing Wc2
followed by ^.dl and sometimes even a4. Or Wei
followed by Jtdl-c2, also maybe combined
with a4. This set-up aims at hindering Black's
infiltration of the light squares with ...Jtd7,
...£>b3 and ...±a4.
8...±d7 9 0-0 (D)
White can also play 9 Sbl first, which may
turn out to have a psychological influence.
Maybe White wants to play 10 b4 cxb3 11 c4!?,
maybe not. Or maybe White just wants to
confuse Black. The safest is probably 9...Wc7, when
White has nothing better than 10 0-0 £\e7 - 9
0-00*710VblWc7.

32
French: Advance and Other Lines
9...£>e7
The most flexible, although there are other
options:
a) 9...0-0-0 10 Sbl <±>b8 11 Wc2 f6!? 12
Sel (D) and now:
B
al) 12...£e8 13 b4 cxb3 14 £>xb3 ±a4 15
£>fd2 fxe5 16 Wa2 £>xb3 17 £>xb3 Wc7 18
dxe5 ±xb3 (18...«xe5? 19 £>d4 ±, threatening
Sxb7+) 19 Wxb3 ±c5 20 ±g4 Se8 21 Af3 ±
Stefansson-Socko, Cappelle la Grande 1998.
a2) 12...£>e7 13 exf6 gxf6 14 b3 cxb3 15
£>xb3 £>xb3?! (15...Aa4 16 Af4+ <±>a8 17 £rfd2
£>g6 18 ±g3 oo) 16 Sxb3 Wc7 17 Ab5!? ±
Ghaem Maghami-Villamayor, Calcutta 2001.
b) 9...f6 10 exf6 (10 Sbl!?; 10 Sel £>e7 11
Sbl Wc7 12 g3 0-0-0 13 Afl ±/= V2-V2 Gal-
dunts-Kiselev, Pardubice 1995; the text-move
has the advantage that it can also be played
against 9...f5) 10...gxf6 11 Sbl ±d6( 11...0-0-0
12 b4 cxb3 13 c4 ±) 12 b3! cxb3 13 c4 dxc4 14
EhxcA £}xc4 15 ^,xc4 ± Jonkman-Apicella,
Cappelle la Grande 2001.
c) 9...J.e7 (this looks rather clumsy at first
sight but Black wants to transfer his knight via
h6 to f5, without having to worry about White
playing g4 in reply to ...£tf5, as this can then be
met by ...£>h4) 10 Sbl ^h6 11 Sel £rf5 12 £rfl
0-0-0 13 Af4 Sdf8 14 £>g3 £>xg3 15 ±xg3 ±
Prie-Apicella, French Ch (Auxerre) 1996.
We now return to 9...£\e7 (D):
10 Sel
This prepares a regrouping of the minor
pieces, usually in the form of £tf 1 and A.f4. An
alternative is 10 Sbl, and then:
W
1H mwm
^Iff m
a) 10...£\b3? loses a pawn to 11 ±xc4!, but
this is surprisingly often missed. Black cannot
recapture as £}xc4 followed by £}d6+ would be
decisive.
b) 10...0-0-0?! 11 £>g5 &e8 12 b4 is much
better for White.
c) 10...£>g6 11 g3 k&l 12 h4 0-0-0 13 h5
£>f8 14 h6 gxh6 15 <±>g2 ± Korchnoi-Khasin,
USSR Ch (Leningrad) 1956.
d) 10...h6 11 £>h4!?(ll Sel -10%el h611
Sbl) 1 l...g5 12 £>hf3 £>g6 13 h3 Ae7 14 £>h2
0-0-0 00.
e) 10...^c7 and now:
el) 11 Sel - 10 %el ^c711 Sbl.
e2) 11 Wei (intending Adl-c2) 1 l...Aa4 12
Adl Axdl 13«xdl£>ec6=.
e3) 11 £>g5 h6 12 £>h3 0-0-0 13 £>f4 g6 (it
is probably a good idea not to allow the knight
to h5) 14 Bel <±>b8 = Qin Kanying-Thoang
Thang Trang, Shenyang worn 2000.
We now return to 10 Bel (D):
I
kmXmkmk
I*il B
^P 4 W& iff 9
J*o^ mm,, Wa
El

Advance: 5..Mb6 6 a3
33
10...H6
Black often elects to move his queen to c7
and then transfer his king's knight to b6 via c8,
so ...h6 is not strictly necessary but it is
frequently a useful move, and often Black plays it
sooner or later anyway. In the game Alekseev-
Iliushkin, Voronezh 1999 Black went without it
for a while: 10...Wc7 11 Sbl £>c8 12&fl £>b6
13 MA £>b3 (13...±a4 is better, as Black would
rather have his bishop as a blockader on b3) 14
£>3d2 MA 15 £>xb3 ±xb3 16 Wcl 0-0-0 17
g4!? Ml 18 £>g3 h6 19 M3 and White might
claim a small edge.
11 Sbl #c7 (D)
Sveshnikov-Casper, Moscow 1987 was
instructive as one of White's main plans was
essayed: 11...0-0-0 12 Wc2 <A>b8 13 ±dl!? (this
set-up aims at preventing Black from
infiltrating with ...±a4-b3) 13...Sc8 14 £>fl £>b3 15
MA <4>a8 16 £>g3 ±a4 17 We2 ®b5 (17...g6 is
possible but White would then try to get a
knight to g4; e.g., 18 h4 followed by £\fl-h2-
g4) 18 £\h5 with an edge for White.
W
m
ahmum
12£>fl
This is clearly White's most common. As the
b2-pawn is defended, White is now ready to
develop his dark-squared bishop. In Kharlov-
Iliushin, Tula 2002 a different plan was chosen:
12 M\ £>c8 13 Se3 £>b6 14 £>el 0-0-0 15 f4
g6 16 Sh3 Ml 17 £>df3 MA 18 We2 and
White may be a bit better but Black remains
solid and may gain counterchances later in the
game. If he survives on the kingside he will
have a future on the queenside as White has
little chance of opening that front but Black
may be able to advance his a- and b-pawns.
12...£ic8 13 ±f4 £>b6 14 ±g3 ±a4 15 Wcl
0-0-0 16 h4 ±b3 (D)
W
The same assessment as in the last note
applies. The game Haba-Naumkin, Cappelle la
Grande 1998 continued 17 £\3d2 (Black was
threatening 17...jta2 followed by ...£ft>3, and
17 #f4? would allow a very favourable switch
of diagonals for Black's bishop with 17...jtc2
followed by ...±g6) 17...±a2 18 Sal ±b3 19
h5 ±e7 20 £>e3 <4>b8 21 Sf 1 <4>a8 22 Wei ±a4
23 ^h 1. White might be very slightly better but
again Black is solid.
D22)
8g3(D)
^8
in*
— 11B
WW W.
y//m m;~mm"""''m
m mmm&
8...i.d7 (D)

34
FRENCH: ADVANCE AND OTHER LINES
It is clear, with Black's little space on the
kingside, that his king is going to be a lot safer
on the queenside, and thus this prepares castling
queenside. However, Black must also make up
his mind around here about how to deal with the
kingside. Interestingly, even now that Black has
moved his queen's knight away from the centre,
this is often combined with playing ...f6 or ...f5.
In a couple of games Korchnoi has chosen
the move-order 8...Jte7 and only after that
played ...Jtd7. It is not clear exactly why this
great master prefers this, but maybe he wants to
cut out White's set-up in the following game:
8...£>e7 9 £>h4!? h6 10 £ig2 ±d7 11 £>e3 0-0-0
12 Wf3 (12 h4!?; 12 ±h3!?) 12...£tf5 13 ±e2
±e7 14 £>xf5 exf5 15 Wxd5 ±e6 (15...±c6 16
Wxf7 Axhl 17 Wxe7 ±) 16 Wg2 g5 17 Wf 1!?
(planning f4, or regrouping with Jtf 3 and tfe2)
17...»c7 18 f4 gxf4 19 Wxf4 h5 20 Ibl h4 21
£>f 1 Idg8 22 Igl Wd7 23 Wf2 hxg3 24 hxg3
Ig7 25 Af4 Igh7 26 Af3 Wb5 27 Idl £>b3 28
Wc2 Wb6 29 £>e3 ± Malaniuk-G.Kuzmin,
Nikolaev Z 1995.
Moreover, after 8...jte7 9 Jth3, Black may
also consider attacking on the kingside straight
away with 9...h5!?.
13 h4 Ic8 14 ±h3 g6 15 Wc2 Ic7 16 £>fl
Wb3 17 We2 ±a4 18 Axf5 gxf5 19 ±g5 gave
White an edge in Nunn-K.Arkell, Port Erin
1994) 11 exf6 (there are no problems if White
doesn't take but now it is clear that White's
position would have appeared more menacing
with the bishop on h3) 1 l...gxf6 12 £>h4 0-0-0
13 Wh5 ±e8 14 We2 £>g6 15 £>df3 Ml 16
±e3 <4>b8 17 Ifel <4>a8 18 ±h3 ±g7 and
Black is at least equal, Golod-Finkel, Israeli Ch
(Ramat Aviv/Modiin) 2000.
b) 9 ±h3 (D) and then:
^mj^m
WV2&.
4 mm 6 flP 4 Wfr 4
1 M «
IB I
1; 'W%^'4*»,
« 4 f&f S ^
w
9h4
This looks more logical than the immediate
9 Jth3 as White's position appears more
harmonious with the pawn on h4. It is also possible
only to develop the bishop to g2:
a) 9 ±g2 £>e7 10 0-0 f5!? (White has a
pleasant position if he is allowed free play on
the kingside; 10...£if5 11 lei ±e7 12 Ibl h5
bl) 9...£>e7 10 0-0 h6 (10...£>g6 is probably
best met by 11 £>el followed by f4) 11 £>h4 (11
£>el leads to the same) 11...0-0-0 12 £>g2 (12
f4!?) 12...<£b8 13 Ibl <A>a8 14 *hl (14 lei
£tec6 15 Jtg4, intending h4-h5 might be an
idea) 14...£>ec6 (14...£>b3?! 15 £>xb3 ±, with
the idea 15...±a4 16 Wf3) 15 f4 g6 16 £>e3 h5 =
Lobron-Hjartarson, Manila IZ 1990.
b2) 9...±e7! 10 0-0 h5 11 We2 0-0-0 (or
ll...g5 12£>elg4 13±g2£>h6 14^2*) 12
£>el g6 13 f4 £>h6 14 £>df3 £>b3 15 Ibl £>xcl
16 Ixcl Idg8 = Movsesian-Radjabov,
Sarajevo 2002.
c) 9 £\h4!?, intending £\g2-e3, is untested as
far as I am aware but might be worth exploring.
9...H5
Or:
a) 9...f5 10 exf6 £>xf6 11 £>e5 ±d6 12 ±h3
±xe5 13 dxe5 £>g8 14 Wg4 g6 15 0-0 £>e7 16
£>f3 £>b3 17 ±e3 Wc7 18 lad 1 was very
comfortable for White in Grishchuk-Apicella,
Bordeaux 2003.

ADVANCE: 5...Wb6 6 a3
35
b) 9...h6!? 10 Ah3 #c6 11 #e2 #a4 12
£tfl &b3 13 Sbl £>e7 14 Ae3 0-0-0 15 £)lh2
£tf5 16 0-0 Ae7 17 h5 Hdg8 18 ±f4 #a5 19
£)g4 #(18 20 *hl #e8 « Vysochin-Andreev,
Barlinek 2003.
c) 9...0-0-0 10 i.h3 (10 &g5 ±e8 11 #f3
might be worth investigating, or 10 Hbl with
the idea 10...f5 11 b3!?cxb3 12 c4) 10...f5! (D)
is Black's main alternative. Then:
W
M
rikmJL
mm, m^n^
■"""if I
• •3 ^
cl) 110-0£>h612£>el£>f7 13£>g2g5was
already better for Black in Klinger-Portisch,
Dubai OL 1986.
c2) 11 exf6 gxf6 12 0-0 h5 13 lei £>h6 14
£>h2! ±e7! (14...£>f7 15 ±g2 was better for
White in Sax-Kuligowski, Warsaw 1979) 15
Wxh5 Idg8 (15...f5!?) 16 #e2 f5! 17 £>df3
£\b3 and now in Grishchuk-Korchnoi, Biel
2001 (where Black had played an earlier... Jte7
than in the move-order given here) White
decided to return the pawn by 18 Jtxh6 Sxh6 19
ladl ±xh4 20 £>xh4 Ixh4 21 ±g2 with
approximately equal play. Instead, 18 Hbl £>xcl
19 Sbxcl f4! gives Black a strong initiative.
10 ±h3 (D)
10...±e7
Here it is not advisable to castle queenside as
after 10...0-0-0 White has the strong reply 11
£>g5!. In Cherniaev-Collins, Port Erin 2001
White was also clearly better after 10...£\e7 11
0-0#c6 12ffe2£ic8 13lel ±e7 14lbl £>b3
15 £ixb3 cxb3 16 £>d2 #b6 17 #dl ±a4 18
Af 1 ±d8 19 ±e2 g6 20 £>f3. 10...£ih6 is a
feasible alternative but usually transposes to our
main line since Black follows up with ... Jte7.
m wfrwm m
110-0£>h612Ibl
Sax-Ree, Amsterdam 1979 gave an
instructive example of the ideas in this line: 12 £tel g6
13 £>g2 0-0-0 14 Ibl <A>b8 15 £if3! #b3 (in
order to reduce the significance of White's space
advantage, Black hopes to exchange queens) 16
We2 ±a4 17 ±g5 ±xg5 18 £ixg5 #c2?!
(possibly already a step in the wrong direction;
18...#b6 looks safer) 19 Wf3! £>b3 (19...#d3
20 #f6! intending £>f4 is clearly better for
White) 20 Wf4! £>d2 21 Ifcl #d3 22 Idl!
£>xbl 23 Ixd3 cxd3 24 £>f3 d2?! (24...£>f5 is
evidently more precise: 25 #cl d2 26 £\xd2
£ixd2 27 #xd2) 25 £>xd2 £>xd2 26 #xd2 £tf5
27 Wg5 ±.
12...0-0-0 13 lei <A>b8 (D)
W
V/X^/s ^^ Y//////j> «^»w '„WJ», ^^ Ovffff.
HI Wfr 9
4k
■
14<£>h2
14 £>fl £>f5 (14...Sde8!?) 15 Ag5 Sde8 16
i.xf5 exf5 17 ±xe7 2xe7 18 £te3 &e6 19 £>g2
1^3 20 #e2 a6 21 £tf4 g6 22 £>g5 with a slight

36
French: Advance and Other Lines
advantage for White, Radulski-Oms Pallise,
Andorra 2002.
14...g6 15 £>df3 Wb3 16 We2 ±a4 17 Ag5
Axg5 18 £>xg5 *c2
Now 19 £>fl Wxe2 20 Sxe2 £>c6 led to
about equal chances in Stevic-Korchnoi,
Leipzig 2002 but from Sax-Ree above we already
know that 19 Wf3! is more testing, although
White's knight stands better on g2 than it does
onh2.
E)
6...±d7 (D)
This makes no real attempt to stop White's
intended plan but as a counter to White's space-
gaining activities on the queenside, Black
intends to take control of the c-file, and keeps the
option of attacking White's pawn-chain if White
advances with b4.
7b4
7 $Le2 is a viable alternative but it is less
consistent with White's plan and abandons the
hope of being able to develop the bishop to d3
directly. Black should probably play 7...£\h6!?
(7...c4 8 £>bd2 £>a5 - 6...c4 7 fobd! foa5 8
k.e2 k.d7) 8 b4 (the attack on b2 is slightly
awkward for White and prompts this anyway in
view of the threat of 8...cxd4 9 cxd4 £\f5; if 8
±xh6 Black has 8...Wxb2!) 8...cxd4, and now
White doesn't necessarily have to recapture
immediately:
a) 9 cxd4 £>f5 (9...Sc8!? 10 ±xh6 gxh6 11
0-0 ±) 10 ±e3 (10 ±b2 £>a5 11 £>bd2 £>c4!? %
Sveshnikov) 10...1.e7 11 0-0 0-0 is given as
slightly better for Black by Sveshnikov. This
isn't much different from Line E2 but it is
definitely an achievement for Black that White
hasn't been able to develop his bishop
immediately on d3.
b) 9 ±xh6 (D) and now:
bl) 9...gxh6 10 cxd4 a5 (I would prefer
10...Sc8 as an attempt to prevent £\c3) 11 b5
£>e7 12 £>c3 a4 13 0-0 Wa5 14 Wd2 £>c8
(14...Sc8? 15 £>xd5! +-) 15 Sfcl £>b6 16 h3 ±
Rogers-Velimirovic, Vrsac 1987.
b2) 9...d3!? 10 ±xd3 gxh6 11 0-0 ±g7 12
Sel 0-0 13 £\bd2 (later Kholmov won a game
with 13 Sa2; then if 13...f6 14 exf6 ±xf6 15
Sae2 White might be able to claim an
advantage in a position that is difficult to assess
though) 13...f6 14 exf6 ±xf6 15 Sa2!? <±>h8
(15...e5!? was suggested by Sveshnikov; e.g.,
16 c4 e4 17 cxd5 exd3 18 dxc6 ±xc6 19 Wb3+
<±>h8 20 Wxd3 Sad8 with compensation thanks
to Black's strong bishop-pair) 16 c4, Svesh-
nikov-Razuvaev, Palma de Mallorca 1989, and
now Sveshnikov gives 16...Sae8 with counter-
play.
c) 9 b5!? £>e7 (9...£>a5 looks like a good
alternative) 10 a4 dxc3 11 £>xc3 £ig6 120-0 £>g4
13 a5 Wc7 14 Ab2 £Mxe5 15 b6 £>xf3+ 16
^,xf3 axb6 was Shabalov-Dolmatov, Linares
open 2000. Now Psakhis suggests 17 axb6 Wb8
18 Sxa8 Wxa8 19 Wd4!?. There is obviously
compensation but whether it is enough for two
pawns isn't immediately clear.
7...cxd4 8cxd4 2c8(D)

ADVANCE: 5..Mb6 6 a3
37
This is important since it prevents White
from developing his knight to c3. 8...£}ge7 is
a eaker and gave White a very pleasant position
m Kunte-Desmukh, Calcutta 2001: 9 £ic3 £tf5
10 £>a4 Wc7 11 ±b2 ±e7 12 Scl a6 13 ±d3
>0 14 0-0 f6 15 Axf5 exf5 16 Sel ±.
Now White has two main options:
El: 9±b2 37
E2: 9±e3 39
These clearly look the most logical as they
prepare the development of the queen's knight
and also retain hopes of being able to develop
the light-squared bishop to d3. Two
alternatives:
a) On 9 ^,e2 Black can just develop with
9...£}ge7 and ...£tf5 but the recommended
approach is 9...a5!? 10 b5 £>xd4! 11 £>xd4 Sxcl
12 Wxcl Wxd4, when Black is indeed doing
very well after 13 Wc3 ±c5 14 0-0 £>e7. Black
is going to get two pawns and a very solid
position for an exchange. Sveshnikov gave 13
&c7 with a perpetual check after 13...Wxal 14
#b8+ <£e7 15 1U6+. Black can try for more
with 13....&C5 after which, for example, 14 Sa2
&xf2+ 15 <£dl Wxg2 16 Sfl ±e7 17 £>c3!
Sfxh2 18 b6 looks very messy.
b) 9 Sa2! ? is a strange move at first sight but
is actually not pointless. The idea is 9...£\ge7
10 Sc2 &f5 11 $Lb2 when White achieves his
ideal set-up after ll...Ae7 12 Jtd3. This idea
needs further tests in practice but turned out
better for White in Mandl-V.Remmler,
Germany tt 1992, when Black played ll...h5 12
±d3 g6 13 0-0 (13 h3!?) 13...±h6. Now 14
£k3! would have been good for White.
El)
9 ±b2 (D)
The bishop doesn't look that well placed here
but White is attempting to maintain a large
possession of space. As opposed to this, the more
natural 9 Ae3 (Line E2) invites Black to
exchange his knight for this bishop by ...£}ge7-f5
or ...£fo6-g4/f5.
9...£>a5
The knight immediately heads for the c4-
square. Often Black is struggling to find enough
room for his pieces if he doesn't do this but a
few other ideas may be worth considering:
a) 9...£>xb4?! 10 axb4 Wxb4+ 11 !ti2 Sc2
was Angelov-Poriazov, Plovdiv 1988 when
Black won quickly after 12 ±c3? Wb3 13 We3
£>h6 14 ±e2? £>f5 15Wd3 2cl+ 16 ±dl Ab5
17 Wxb5+ Wxb5 18 <S?d2 2xdl+ 0-1. Much
stronger is 12 ^,a3!, as pointed out by Parma.
Now, for example, 12...1rb3 13 Wd3 ±b4+ 14
£>bd2 Wc3 15 Wxc3 ±xc3 16 Sbl Sa2 17 ±d6
is clearly better for White.
b) 9...£>ge7 10 £>c3 £>a5 (10...£>f5 11 £>a4
Wd8 12 ±d3 ±) 11 £>a4 Wc6 12 £>c5 £>c4 13
Wb3 b5 14 Ad3 (14 a4!7) 14...Sb8 15 0-0 £>c8
16 ^.cl looked in White's favour in F.Roder-
Hertneck, Marktredwitz 1984.
c) 9...£h6!?10£c3(10.&.d3?!£>a5 110-0
£>c4 12 ±xc4 Sxc4 13 £>c3 Wa6 was
comfortable for Black in Korchnoi-Kotsur, Istanbul OL

38
FRENCH: ADVANCE AND OTHER LINES
2000) 10...£>a5 11 £>a4 ®c6 12 Scl £>c4 13
Jtxc4 dxc4 14 £\c3, intending d5 with an edge
for White is given by Korchnoi but Black looks
alright after 14...b5 15 0-0 Wb7 16 d5 exd5 17
£\d4 (Olesen-Whiteley, Newcastle 1995), when
17...£rf5 is fine for Black. 14...Wt>6 intending
...a5 is another idea for Black.
10 £>bd2
10 ±d3 £>c4 11 ±c3 £>h6 is fine for Black.
10...£ic4 (D)
11 £>xc4
Even though 11 Jtxc4 dxc4 12 Scl seems
to speed up White's development, some
trouble awaits him after 12...c3! (12...Wa6 13^e4
±c6 14 £>fd2 ±) 13 Sxc3 Sxc3 14 ±xc3 Wa6.
Since White is unable to castle for the
moment, Black had satisfactory compensation in
Orlov-Rotsagov, St Petersburg 1995 after 15
£>e4 ±a4 16 Ifbl ±b5, and 15 ±b2 £>e7 16
£\g5 h6 17 £\ge4 £\d5 also looked fine for
Black in U.Voigt-Touzane, Biel 1995.
Il...dxc4 12 Scl a5
This is the most common but not necessarily
Black's best. 12.. J?a6!? (D) is the alternative:
a) 13 d5 exd5 14 Wxd5 had been considered
better for White on the basis of Afek-Psakhis,
Israel 1990, in which 14...±e6 15 We4 was
played, but since then it seems to have been
established that Black should delay the blockade
of White's e-pawn, and instead try 14...£te7 15
We4 b5 16 ±e2 Wg6, which indeed looks fully
playable for Black.
b) Perhaps in view of this, White tried 13
£\g5!? in Peng Xiaomin-Cheng Xinkal, Suzhou
2001. After 13...b5 14 Wh5 £>h6 15 Bc3 White
went on to win a fine attacking game but if Black
just continues 15... Jtc6! followed by ...Wbl and
maybe ...a5,1 don't see why this should be so
fantastic for White.
13 £>d2 axb4 14 ^xc4 Wd8 (D)
Black can prevent White's next move with
14...Wa7?! but the queen is then on an inferior
square. 15 axb4 and now:
a) 15...±xb4+ 16 ±c3 ±xc3+ (16...±e7
may be better, but 17 d5 exd5 18 #xd5 is good
for White) 17 Sxc3 *f8 18 Sa3 WbS 19 £>d6
Sd8 20 »f3 £>h6 21 ±d3 was very good for
White in Sveshnikov-Nevednichy, Bled 1991.
b) 15...b5 16 £>d6+ ±xd6 17 2xc8+ ±xc8
18 ±xb5+ ±d7 19 ±xd7+ Wxd7 20 exd6 £>f6
21 0-0 0-0 22 Wb3 Sb8 left Black with some
chances to defend in Fressinet-Harikrishna,
Calicut jr Wch 1998, but with a pawn less even
after regaining d6, this is obviously an uphill
struggle.

ADVANCE: 5..Mb6 6 a3
39
15 a4!?
This might well be an improvement on 15
axb4 ±xb4+ (15...b5 16 £>d6+ ±xd6 17 exd6
£>f6 18 ±d3 0-0 19 0-0 ±c6 20 d5! is better for
White according to Sveshnikov) 16 Jtc3 Jtxc3+
17 2xc3 which was initially assessed clearly in
White's favour but 17...4>f8! has turned out to
be a stubborn defence. Black is temporarily
passively placed but there are sound positional
prospects when the bishop comes to c6 and the
knight joins the game. We need to look more
critically at this, of course, and currently the
main line runs like this: 18 Wal ±c6 19 £>d6
Ia8 20 la3 Ixa3 (20...b6 attempts to defend
the a-file but 21 Jtb5 could be good for White,
with the idea 21...±xg2 22 Igl ±d5 23 Igg3!?
Ixa3 24 Wxa3 £>e7 25 Ic3 ±/+-) 21 Wxa3
£te7, and now 22 h4!? looks like the critical
test. Following 22...h5 23 Ih3 g6 24 Ib3 4>g7
25 Wc5 ±d5 26 ±c4!? ±xc4 (26...±xg2!?) 27
Sxb7, White won easily in Reefat-Bengtsson,
Stratton Mountain 2000 after 27...£>c8? 28
£>xf7, but 27...Wa8! might hold.
15...±c6 16 a5! (D)
With this move White creates the possibility
of obtaining an outpost on b6. In general White
should attempt to exploit his slight initiative
because Black is doing well from a structural
point of view but is a little behind in
development. It remains to be shown whether Black can
neutralize White's pressure:
a) 16...£>h6 17 £>b6 Ib8 (17...Ic7 18 d5
exd5 19 ±d3 ±e7 20 0-0 0-021 ±bl also gives
White a very strong attack) 18 d5 (18 ±c4!?)
18...exd5 19 ±d4?! (19 ±d3! ±e7 20 0-0 0-0
21 Jtd4 intending a6 was very good for White
in Dm.Schneider-Whatley, Buenos Aires 2003)
19...£if5? (19...b3!, intending 20 Wxb3 #h4,
gives Black excellent counterplay) 20 a6 Jte7
(20...b3 21 a7 ±) 21 a7 Ia8 22 £>xa8 Wxa8 23
±b6 +- Skytte-Vovk, Danish Ch (Arhus) 1999.
b) 16...Ic7 17 Wb3 Wd5 18 f3 £>h6 19 ±d3
£>f5 20 ±e4 Wd8 21 Axf5 exf5 22 £>b6 Wh4+
23 g3 Wh3 24 *f2 ± Fressinet-Shirov,
Bordeaux rpd 1999.
c) 16...±e7 17£>b6lc718d5exd5 19±d3
±g5 20 Ic2 £>e7 21 0-0 ±h6 22 ±d4 0-0 23
Wbl and White is better, Fressinet-Socko, Bun-
desliga 2002/3.
d) 16...Ia8!? 17 ±e2?! (this move looks
slightly passive; better is 17 a6 £\h6 18 £te3
Wdl 19 Wc2 ±a4 20 axb7 Wxb7 21 Wc7, or 17
lal!? £>h6 18 £>b6 Ia7 19 Wc2, maybe with a
slight edge for White in both lines) 17...£\h6
(17..Jtxg2 18 Igl ±c6 19 a6 with good
compensation) 18 ±f3 ±el 19 0-0 0-0 20 ±xc6
bxc6 ? Abram-Matlak, corr. 1996.
E2)
9 ±e3 (D)
9...£>h6
The most common route for Black's knight
to threaten White's dark-squared bishop is via
h6 to g4. On 9...4ftge7, Black must reckon with
10 ±d3 £>f5 11 ±xf5 (110-0 £>xe3 - 9...*hh6
10 ±d3 £hg411 0-0 £hxe3) 11 ...exf5, leading to
a type of position that occurs in several lines of

40
French: Advance and Other Lines
the Advance Variation. White might have a
very slight edge due to his space advantage but
Black's position remains very solid. 12 0-0 (12
£>c3? £>xb4) 12...±e7 13 £>c3 (D), and now:
a) 13...^xb4 is possible but should lead to
White's advantage; e.g., 14 Scl (14 £>a4 ±xa4
15 Wxa4+ £>c6 16 Sfbl Wa5 17 Wc2 0-0 18
Wxf5 ±; 14 Wb3!?) 14...^c6 15 £>xd5 Wa6 16
£>xe7 £>xe7 17 2xc8+ ±xc8 18 Wcl 0-0 19
Wc5!±
b) 13...Ae6 14 £>a4 Wd8 15 Wd2!? (a
normal plan for Black would be to seize space on
the kingside with ...g5, so White attempts to
stop this before moving in with the knight on
c5) 15...h6 (15...b6!?) 16 £>el 0-0 17 £>d3 <&h7
18 £>ac5 b6 19 £>b3 Wd7 20 £>b2 £>b8 21 Wd3
± Jonkman-Stevanovic, Lisbon 2000.
10 ±d3 (D)
10...£>g4
10...£te7!? is an alternative plan that has
done quite well. Black prepares to put a knight
on f5 without being obliged to have his pawn-
structure weakened. Besides, Black hopes to
exchange bishops with ...^.b5. 11 ^.xh6
(usually this exchange is nothing to be afraid of but
here it might be White's best; 11 We2 £>ef5 12
0-0 ±e7 13 h3 £>xe3 14 fxe3 £>f5 was fine for
Black in Giaccio-An.Rodriguez, Mar del Plata
2001) ll...gxh6 12 0-0 (12 Sa2 followed by
Sc2 before castling might be more accurate)
12...^.b5 (and here Black should probably play
12.. JLg7 with the idea 13 Sa2 0-0 and White
cannot play 14 Sc2 in view of 14..JLa4) 13
Sa2 ±g7 14 Sc2 0-0 15 Sc5, and White was
better in Romero-Grau Dominguez, Cullera
2001.
11 0-0 ±el (D)
ll...£}xe3 12 fxe3 g6 is similar to the line
6..&h6 7 b4 cxd4 8 cxd4 *hf5 9 Ae3 thxe3 10
fice3 g6, and it is worth comparing the two. The
game Jonkman-Carrasco Martinez, Lisbon 2001
continued: 13 Wd2 (13 £>g5? £>xe5!) 13...±h6
(13...±g7!?) 14 g4!? a5?! (14...Wd8!? could be
considered, with the idea 15 h4? £>xd4!) 15 b5
£>d8 16 a4 Af8 17 Sa2 ±e7 18 g5 Wa7 19 h4 h5
20^el±.
12Sa2!?
This interesting attacking concept has
become White's most intimidating set-up. The
little rook move is played in anticipation of an
exchange on e3, after which the rook may swing
to the f-file, thus significantly increasing the
power of a kingside attack. Other options are:

ADVANCE: 5..Mb6 6 a3
41
a) 12 Wd2 £>xe3 13 fxe3 - 6...Zhh6 7 b4
cxd4 8 cxd4 Zhf5 9 ±e3 kd710 ±d3 *hxe3 11
fxe3 kel 12 0-0 Zc8 13 ^d2.
b) 12 £>bd2 £>xe3 13 fxe3 (D) and now:
bl) 13...0-0 14 £>b3 a5 (14...^b8 15 We2
a6 16 £>c5 Ab5 17 i.xb5 axb5 ± Torre) 15
£>c5 JLe8 16 £>a4 Wd8 17 b5 £>b8 18 Sa2
£>d7 19 Saf2 £>b6 20 £>xb6 Wxb6 21 a4 Sc3!
*> Miljanic-Antic, Yugoslavia 1997.
b2) 13...^b8!?14^b3(14*e2Aa4 15h4
0-0 16 Sf2 Wd8 17 £>fl!? i.xh4 18 g3 Ae7 19
5Mh2 £id7 20 Saf 1 gave White compensation
in Otero-Pecorelli, Cuba 1999) 14...i.a4 15
ffbl i.xb3 (15...Sc3 16 £>c5 i.xc5 17 dxc5
#d8 =) 16 Wxb3 0-0 17 Sacl g6 = Kunte-
H.Gretarsson, Elista OL 1998.
We now return to 12 Sa2!? (D):
12...£ixe3
12...0-0 13i.f4!?f6 14exf6£>xf6 15Sc21ed
to a quick disaster in Olofsson-J.Fries Nielsen,
corr. 1998 after 15...<&h8?! (15...a6 is probably
better; e.g., 16 Ae3 i.d6 17 £>e5 &e8 with a
fairly normal position) 16 JLe3 a6 (16...JLd6 17
£>e5 is also better for White; e.g., 17...a5 18 b5
£>xd4? 19 £>xd7 £>xd7 20 Wg4 Ae5 21 Wh5
+-) 17 £>e5 Ae8 18 Wf3 ± £>a7 19 Sxc8 £>xc8
20 Wh3 <&g8 21 Scl Ah5?? 22 Sc6! i.xd3 23
Wxe6+ 1-0.
13 fxe3 0-014 Saf 2 Wd815 £>bd2 g616 h3
f5 17 exf6 Bxf6 18 b5 £>a5 19 £>e5
White has an advantage, Grabarczyk-Szy-
manski, Koszalin 1998.

2 6±e2
Ie4e6 2d4d5 3e5c5 4c3&c6 5£>f3lrb6 6
±e2 (D)
With this White continues to develop
sensibly but it lacks the aggression of 6 a3, and
Black can now more easily step up the attack
against d4 by transferring a knight to f5.
The two main choices are:
A: 6...£>h6!? 42
B: 6...cxd4 44
Minor options include:
a) 6..±dl-5..±d7 6k,e2^b6.
b) 6.. .£}ge7 7 dxc5 (a typical way of trying to
punish Black for omitting the exchange on d4,
but in this case not offering White anything; 7
£>a3 £>f5 8 £>c2 cxd4 9 cxd4 - 6...cxd4 7 cxd4
foge78tha3*hf59thc2)l..Mxc5 (7...Wc7?! 8
£>d4! ±) 8 Af4 (8 Ad3!? £>g6 9 #e2 d4 10
±xg6 hxg6 oo; 8 £>a3 £)g6 9 £>b5 #b6 10 ±e3
#d8 11 0-0 oo) 8...£>g6 9 Ag3 Wb6! 10 b4 ±d7
11 ±d3 a5 12 b5 £>ce7 13 £>bd2 £>f5 14 We2
a4!? with good counterplay, Benjamin-Korch-
noi, Horgen 1994.
c) 6...f6 7 0-0 cxd4 8 cxd4 fxe5 9 dxe5 (the
usual advice of recapturing with the knight first
does not apply here, since 9 £}xe5 can be met
by 9...#xd4) 9...g6 (there may be better moves
here, but White has the advantage anyway) 10
£>c3 ±g7 11 ±e3 Wd8 12 £>b5 is much better
for White, Ganguly-Satyapragyan, Goa jr Wch
2002.
A)
6...£>h6!? (D)
W
tip a gp^u^
iHJLil
p n# kW§ §
Leaving out the exchange on d4 has the
advantage of avoiding Line B2, where White has
the option of playing his knight to c3.
7±xh6
This is critical. Instead 7 b3 cxd4 8 cxd4
transposes to 6...cxd4 7 cxd4 £hh6 8 b3 (Line
Bl), and 7 £>a3 £tf5 8 £ic2 cxd4 9 cxd4 to
6..xxd4 7 cxd4 Zhge7 8 *ha3 £>/5 9 Zhc2.
Benjamin seems to favour 7 ^.d3, which can
be compared with the line 5...^hh6 6 &d3, only
here Black has his queen on b6. It isn't clear
whom this change favours. After 7...cxd4 8
cxd4 there are two options for Black:
a) 8...£>f5 9 ±xf5 exf5 10 £>c3 ±e6 11 0-0
h6 12 £>a4 #b5 13 £>el! (preparing ±e3 and
£kI3, and intending to answer ...g5 with f4)
13...±e7 14 £e3 Ic8 15 £>d3 b6?! (Timman
queries this and suggests 15...0-0 instead) 16
£>c3#c4(16...#a6 17«tf4±) 17 Scl g5 18 f4

6 ±e2
43
g4 19 a4! £)b4 20 £)f2! and White is better,
Timman-Andersson, Malmo 2000.
b) 8...^.d7 (this is how Black plays in the
Milner-Barry Gambit, so why not here where
Black has ...£)h6 as an extra move, good or
bad?) 9 ±c2 (9 0-0 £ixd4 10 £ixd4 #xd4 11
£)c3!?) 9..M5 (9...f6!? - J.Watson, or 9...£)b4
=) 10 ±xf5 exf5 11 £ic3 ±b4!? (Il...±e6 -
8...Zhf5 9 ±xf5 exf510 &c3 ±e6) 12 *f 1!? 0-0
13 *gl ±xc3 14 bxc3 Ifc8 15 h4 £)d8 16 ±d2
Sc4 looked fine for Black in Benjamin-H.Gre-
tarsson, Bermuda 1999.
We now return to 7 ±xh6 (D):
7...gxh6
7...#xb2? would be possible if Black had
exchanged on d4 before playing ...£)h6, but
here it is a fatal error in view of 8 ^.e3! #xal 9
#c2 cxd4 10 £ixd4 £ixd4 (10...±a3 11 £*5!)
11 ±xd4 ±a3 12 ±b5+ *f8 13 0-0, when
White wins because 13...#b2 is met by 14
±c5+.
8 #d2 ±g7 9 0-0 0-0 10 ^a3 cxd4
The insertion of this exchange is advisable
since 10...f6?! 11 exf6 Ixf6 gives White the
option of playing 12 dxc5! #xc5 13 b4 #f8 14
£ic2 ±d7 15 b5 £ie7 16 £)e5 ±e8 17 £)d4 with
the better position for White, Kupreichik-Khuz-
man, Sverdlovsk 1987.
Another idea is 10..JLd7 11 £ic2lac8 12b4
(12 dxc5!?) 12...cxd4 13 cxd4 £ie7 14 a4 #c7
15 Ifcl #c3 16 #dl ^g6, as in Rajlich-
Lputian, New York 1998. Now Psakhis
suggests 17 ^e3!? #xb4 18 Icbl #a5 19 Ixb7
i.c6 20lbbl^f4 21i.b5±
11 cxd4 f6 12 exf6 lxf6 13 &c2 (D)
White is structurally better but Black has a
very active position and some straightforward
plans in the form of... JLd7-e8-g6/h5, or an
exchange sacrifice on f3. White counters this by
advancing his b-pawn, intending to dislodge
the black knight from controlling e5.
13...±d7
Black might also take measures against
White's plan with 13...a5 14 £ie3 ±d7 15 £)g4,
and now:
a) 15...Ixf3 16 Jbtf3 #xd4 17 ^xh6+ (17
ladl ±) 17...*h8 18 £tf7+ 4>g8 19 ^h6+
<A>h8 20 £)f7+ (20 #g5 <*,) 20...4>g8 21 £)h6+
&h$ V2-V2 Schandorff-Antonsen, Copenhagen
1995.
b) 15...Ig6 16 ^ge5 ^xe5 (16...^xd4 17
^xd7 ^xf3+ 18 Axf3 #c7 19 ±h5 Ig5 20 g4
#xd7 21 f4 Ixh5 22 gxh5 ±) 17 £ixe5 ±xe5 18
dxe5 ±.
14 b4 (D)
Here the plan with £)e3-g4 is probably too
slow due to 14 £)e3 Saf8 and now 15 £)g4 can
be met by 15...Sf4.
14...Iaf8
This is the most common but other options
are also worth considering:
a) 14...£ie7 15 £ie5 i.a4 16 b5!? Ic8 worked
well for Black in D.Howell-Kelly, Bunratty
2001 after 17 £ie3?! If4! ?, but 17 £)d7! Ixc2
18 #b4 would have cast doubt on Black's
idea, in view of 18...#d8 19 £)xf6+ ±xf6 20
±d3 ±, or 18...Ixf2 19 Ixf2 #xd4 20 #xe7
and White wins.

44
French: Advance and Other Lines
b) 14...Exf3!? 15 ±xf3 £>xd4 16 £>xd4
Wxd4 17 Wxd4 ±xd4 18 Sacl ±b6 held
surprisingly easily for Black in Hort-Spassky,
Moscow 1999 but objectively White is pressing.
c) The above idea can also be prefaced by
14...a6 15 a4 and only then 15...Sxf3!? but
again I doubt that the sacrifice is 100% correct.
Instead in Savic-Antic, Yugoslav Ch (Banja
Kovijaca) 2002 Black continued more
cautiously with 15...Sff8 but White was better
after 16 Sa3 ±e8 17 Se3 £>e7 18 a5 !tt6 19
±d3.
d) 14...±e8!? 15 b5 £>d8 16 a4 (16 £>e5!?)
16...£>f7 17 £>e3 h5 18 Sacl £>d6 19 Wb4£>e4
20 a5 ©d8 <*> Fernando-Gdanski, Cappelle la
Grande 2002.
We now return to 14...Baf8 (D):
W
15 b5 £>e7
This is by far the most common move but
15...£}a5!? might not be bad. Much depends on
the assessment of 16 £ie5 ±xb5 17 Sabl ±xe2
18 Sxb6 ±xfl 19 Sxe6! (in Jonkman-Tiggel-
man, Vlissingen 1999 White had nothing after
19 Sbl Sxf2 20 Wxf2 Sxf2 21 <&xf2 ±a6)
19...Sxe6 20 Wxa5 ±c4 21 £>xc4 dxc4 22 1U5
Sfe8 23 ©xc4 &h8, which I am not sure about,
but possibly is it better for White.
16 £>e5 ±e8 (D)
w
17 g3
This slightly weakens the kingside but White
hopes to be able to support his e5 stronghold
with a later f4. The move also aims at
discouraging ...£}g6, which turned out well for Black
in Adams-Lobron, Amsterdam 1994 after 17 a4
£>g6! 18 £>g4 S6f7 19 £>xh6+ ±xh6 20 ©xh6
£rf4, with strong counterplay. An interesting
idea which needs further investigation is 17
^.d3!? £}g6 (perhaps Black needs something
else here) 18 ±xg6 hxg6 19 a4 Ef5 20 Sael g5
21 £>d3 a6 22 ^c5 ± Thipsay-Konguvel,
London 2001.
17...h5
Black's ...£}g6 plan is now less effective:
17...£>g6 18 £>g4 S6f7 19 £>xh6+ ±xh6 20
©xh6 ±xb5 21 Sabl Wc7 22 Sxb5 ©xc2 23
«e3±
18 a4 £>f5 19 a5 !fc7 20 Sacl £>d6 21 £>e3
«e7
Black has counterplay, Kupreichik-Lautier,
Belgrade 1988.
B)
6...cxd4 7 cxd4 (D)

6 ±e2
45
7...£>h6
Or:
a) 7...±b4+ 8 £>c3 £>ge7 9 0-0 ±xc3 10
bxc3 ±d7 11 ±d3 £>g6 12 h4 ± Keitlinghaus-
Krukowski, Porz 1989.
b) 7...^ge7 gives White the possibility of
transferring the queen's knight via a3 to c2: 8
£>a3 £>f5 9 £>c2. This is probably still a slightly
better version for Black of the line 5...£foge7 6
£ha3 cxd4 7 cxd4 £l/5 8 *hc2 ^b6 where White
can play 9 k.d3. After 9 £>c2, Black has:
bl) 9...£>b4 10£>e3!?±
b2) 9...±b4+ 10 *f 1 ±e7 11 h4 h5 12 b4!?
(12 g3!? ± Bareev) 12...±d7 13 a3 Sc8 14 g3
£>d8 (14...£>b8!? with the idea of ...±b5 «
Bareev) 15 £>e3 ±b5 16 ±xb5+ Wxb5+ 17
*g2 Sc3 18 Sel £>xe3+ 19 Sxe3 Sc4 20 Sd3
£>c6 21 ±e3 ± Benjamin-Bareev, Munich 1994.
We return to 7...£>h6 (D):
Now:
Bl: 8b3 45
B2: 8£>c3 46
Note that here, in the event of 8 ^.xh6?,
Black can safely capture on b2.
Bl)
8 b3 £>f5 9 ±b2 (D)
In similar positions White often has a second
choice in JLe3 but here, 9 3Le3 makes little
sense, since Black can reply 9.. JLb4+. Then 10
if 1 £>xe3+ isn't attractive for White, and after
10 £ibd2 ±c3 he loses the d4-pawn.
9...±b4+
White loses the right to castle after this but his
king is not that badly placed on f 1 as it can easily
be improved after, for example, g3 and ig2.
Meanwhile Black should also be prepared to
meet a white g4. Another option is 9.. JLd7, and
White must then dislodge the black knight from
f5 in order to develop his queen's knight:
a) 100-0 Ae7 11g4(ll»d2 0-0 12Sdlf6
with good counterplay) ll...£ih4 12 £lxh4
±xh4 13 £>a3 f6! 14 exf6 ±xf6 15 £>c2 0-0 ?
Mortensen-Cu.Hansen, Danish Ch (Aalborg)
2000.
b) 10 g4!? £>fe7 (10...£>h6 11 h3 <*>) 11 £>c3
h5! 12 £>a4 Wd8 13 g5 £>g6 14 #d2 Sc8 15
Scl ±b4 16 ±c3 ±a3 = Kupreichik-Korchnoi,
Senden 1996.
A second alternative is 9...^.e7 but I find it
hard to believe that it is better than 9...k.b4+ 10
*/7 SLe7.

46
FRENCH: ADVANCE AND OTHER LINES
10 4>fl (D)
10...±d7!?
Black's main idea is to castle and break up
White's centre with ...f6, but this move tries to
make a white £>c3 less favourable.
Alternatives:
a) 10...0-0 11 £ic3 (nowadays White tends to
reject 11 g4 £>h6 12 Igl on account of 12...f6!
13 exf6 Ixf6 14 g5 Ixf3 15 Axf3 «M5 with
good counterplay for Black - compare with the
mainline) ll...f6 12£>a4#c7 (12...»d8!?) 13
g4 £>h6 (13...fxe5 14 gxf5 e4 15 £>e5 ±;
13...£tfe7 14 a3 £a5 15 b4 £b6 16 b5 £>a5 17
Scl ±) 14 exf6 (14 a3 looks more accurate,
although the position remains very complicated)
14...gxf6 (14...£ixg4 15 fxg7 Sf7 is very strong)
15 Sgl Jtd6 16 JLd3 with a very complicated
position, Warmbier-Polklasser, corr. 1995.
b) 10..JLe7 and then:
bl) 11 h4 #d8!? (to prevent g4, but 1 l...h5 is
also possible, or ll..JLd7) 12 h5 f6 13 a3 fxe5
14 dxe5 0-0 15 £d3 ±dl 16 £>bd2 Wb6 17 b4
£>cd4 18 *gl £>xf3+ 19 £>xf3 a5, Movsesian-
Priehoda, Slovakian Ch (Kaskady) 2002.
Despite the loss of tempi with the queen, Black has
excellent counterplay.
b2) 11 £>c3 Wd8 (1 l...f6!? is possibly more
accurate, with the idea 12 g4 £ih6 13 exf6 gxf6
14 #d2 fan with counterplay) 12 g3 f6 13
*g2 fxe5 14 dxe5 0-0 15 Icl ±dl 16£d3!? ±
Sveshnikov-Portisch, Biel IZ 1993.
11 g4
11 £>c3 £>cxd4 12 £>xd5 exd5 13 £>xd4
£>xd4 14 !,xd4 JLc5 is equal.
The text-move is critical but appears to
permit Black good counterplay.
U...&h612 Igl 0-013 a3 ±e714 b4 f6! 15
exf6 Sxf6 (D)
Damljanovic-Velimirovic, Yugoslav Ch
playoff (Belgrade) 1997. Black has excellent
counterplay. The critical line appears to be 16 g5
Ixf3 17 £xf3 £>f5 18 Sg4 «, as in Wester-
inen-Antonsen, Torshavn 1997.
B2)
8 £k3 «M5 (D)
9£>a4
This defends the d-pawn by forcing Black's
queen to relinquish its pressure on d4. 9 Atfl?
has been played a number of times, initially
thinking that White defends the d-pawn
tactically but this seems not to be the case, because

6 ±e2
47
9...£>fxd4! 10 ^a4 (or 10 ±e3 #xb2 11 ^xd5
#xe2+! 12 #xe2 ^xe2 13 &c7+ 4>d7 14 ^xa8
£>c3 -+) 10...1rb4 11 ±d2 #e7 12 ±g5 f6 13
exf6 gxf6 14 £ixd4 fxg5 15 ±h5+ 4>d8 simply
wins a pawn, and White's attack doesn't look
very convincing, as shown by several games.
9...#a5+
Often Black's move-order is 9.. JLb4+, when
10 JLd2 #a5 leads to the main line, but then
Black must also reckon with 10 4*1 #d8 11
h4!?. Although not such a big worry for Black, I
see no reason for Black to give White this
option.
10£d2
10 *f 1 b5 is fine for Black.
10...i.b4 11 ±c3 (D)
ll...a6!?
An interesting idea, which isn't new but has
been taken up again recently. Black's main idea
is that Black's queen has an extra square in the
event of 12 a3? ±xc3+ 13 ^xc3 #b6 14 ^a4
#a7, and White cannot defend d4. These
alternatives for Black aren't bad either:
a) ll...Axc3+ 12 £>xc3 #b6 13 Ab5 Ad7
14 ±xc6 ±xc6 15 #d2 ±b5 16 ^xb5 #xb5 =
Shaw-Bryson, Scottish Ch (Aviemore) 1997.
b) ll...b5!? 12 a3 ±xc3+ 13 £)xc3 b4 14
axb4 #xb4 (D) and now:
bl) 15 #a4 ±d7 16 #xb4 (16 ±b5 - 75
iM k.d716 Wa4) 16...£>xb4 17 4>d2 4>e7 18
g4 £>h6 19 £iel Ihb8 20 4>e3 f6 21 f4 fxe5 22
fxe5 ±e8 = Nemec-Prandstetter, Trnava 1980.
b2) 15 ±b5 ±d7 16 #a4 (16 ±xc6 ±xc6
17 #d2 £b5 18 £ixb5 #xb5 19 Ia5 #b6 was
equal in Nunn-Schmittdiel, Dortmund 1991)
16...#xb2!? 17 £ixd5!? exd5 18 0-0 £)fxd4
(18...^b8? 19 Axd7+ £>xd7 20 Ifbl #e2 21
Hb7 Id8 22 e6 +-) 19 Ifbl £>xf3+ 20 gxf3,
Kupreichik-Ulybin, Moscow 1989, and now the
simplest is 20...#d4 21 ±xc6 #xa4 22 ±xa4
±xa4 23 Ixa4 0-0 =.
12 g4 £>fe7 13 0-0 Axc3 14 ^xc3 f6 15
#d2 (D)
15...0-0
15...fxe5 16 £ixe5 £)xe5 17 dxe5 £>c6 18 f4
^.d7 <*> Jonkman-Barsov, Antwerp 1999.
16 ±d3 ±d7 17 lael fxe5 18 £ixe5 £ixe5
19 Ixe5 If 7
with chances for both sides, Jonkman-Kas-
imdzhanov, Wijk aan Zee 1999.

3 6 ±d3 and the Milner-Barry Gambit
Ie4e62d4d53e5c54c3£k65&f3®b66
±d3 (D)
In his book Four Gambits to Beat the French,
Tim Harding promotes this as the 'Murderous
Milner-Barry'. The British master Sir Stuart
Milner-Barry developed the idea of sacrificing
the two centre pawns for a large lead in
development in the early 1950s, but to be honest, I
have little faith in this line if Black plays
correctly.
Black must, of course, avoid the immediate
capture due to 6...cxd4 7 cxd4 £ixd4?? 8 £ixd4
0xd4?? 9 ±b5+. The correct move-order is:
6...cxd4
6.. JLd7 has a point in not releasing the
central tension yet, but then White need not follow
the main line, and can try 7 dxc5 ^.xc5:
a) 8 b4? (an enterprising idea, but not
correct) 8...±xf2+ 9 &e2 £>xb4! 10 cxb4 ±d4 11
£>a3 (11 £>xd4 «xd4 12 «b3 «xe5+ 13 <&dl
£>f6! 14 ±b2 «g5 is much better for Black)
ll..Jk.xal 12 Ae3, T.Clarke-Roberts, Irish Ch
1995, and now 12...d4! 13 «xal dxe3 14 £>c4
0xb4 would have put White's idea into
question.
b) 8 0-0 a5 (it is best to prevent White from
gaining space with b4) 9 0e2 f6!? (9...£>ge7 10
MA £>g6 11 ±g3 0-0 followed by ...f6 also
seemed OK for Black in Nun-Kupreichik,
Germany tt 1991/2) 10 c4 £>ge7 11 £>c3 (11 exf6
gxf6 12 £>c3 £>b4 13 cxd5 £>exd5 14 £>xd5
£>xd5 oo) 1 l...£ixe5 12 £>xe5 fxe5 13 cxd5 exd5
14 0xe5 0-0 with counterplay, Nun-Repkova,
Pardubice 1992.
7 cxd4 ±d7 8 0-0
We shall devote some time to discussing the
move-order around here since 8 £ic3 is also an
option. 8 ±e2 and 8 ^.c2 are also seen but are
not especially good. The first of these loses a
tempo on well-known lines while the latter is
well met by 8...£ib4, gaining the bishop-pair.
White's main argument for choosing 8 £ic3
over 8 0-0 is that it avoids 8...£>b4 but that
move loses lots of time, and despite it gaining
the bishop-pair, Black rarely opts for it. White
also loses some options later on if he chooses 8
£)c3 but usually it merely transposes to lines
covered later; e.g., 8...£>xd4 9 £>xd4 0xd4 and
now 10 0-0 transposes to 8 0-0 *hxd4 9 *hxd4
Wxd4 10 Zhc3, and 10 «e2 £>e7 (10...±b4!?)
11 0-0 to 8 0-0 thxd4 9 ^xd4 ^xd4 10 We2
£he71l£hc3.
8...£>xd4 (D)
White has already won a moral victory if
Black doesn't take the pawn.
W
*m
n Iio H

6 k,d3 AND THE MlLNER-BARRY GAMBIT
49
Now we look at:
A: 9£)g5 49
B: 9£ixd4 50
9 £)bd2 is another gambit, with which the
Czech IM Jiri Nun has scored quite well, and it
is possibly not worse than any of White's other
moves but objectively, and as I have said earlier,
I find it difficult to believe in the whole line for
White:
a) 9...&e7 10 ^xd4 ©xd4 11 &f3 Wa4 12
b3 Wa5 13 ±d2 Wd8 14 Sel gave White ample
compensation in Nun-Razuvaev, Sochi 1989.
b) 9...£)c6 10 £)b3 £)ge7 11 ±e3 Wc7 12
Scl £)g6 13 £k5 (D) and then:
bl) 13...±xc5 14 ±xc5 £)gxe5 15 Sel (in
the recent game Pap-Kosic, Bosnia 2005 White
proved that even 15 £ixe5!? Wxe5 16 Wb3 b6
17 ±a3 Wf4 18 Sfel!? is not without venom)
15...Sc8andnow 16±a3£ixd3 17Wxd3£ie5!
18 Sxc7 ^xd3 19 Bxc8+ ±xc8 20 Sdl &f4
gave Black a won endgame in Nun-Schmittdiel,
Prague 1990 but Nun was prepared to play this
line again in a later game against Brodsky, so
we must assume he had some improvement in
mind. Maybe 16 Af 1!?, with the idea 16...£>xf3+
17 «xf3 ^e5 18 Wg3 ^g6 19 Wxc7 Sxc7 20
JLxa7 and despite the missing pawn, White is
not doing badly because of his two vs one
majority on the queenside.
b2) Maybe Black shouldn't give up his dark-
squared bishop. In Baron-Matamoros, Malaga
2005 Black escaped from 13...±e7!? 14 b4
#b8 15 Sel ±c8 16 b5 &cxe5 17 ^xe5 ^xe5
18 Af4 ±d6 19 ±xe5 ±xe5 20 «h5 Af4 21
Bc2 g6 22 Wxd5 0-0 23 ©B ±xh2+ 24 *hl
^.d6 with a safe material advantage.
A)
9 £g5 (D)
This has been especially popular in
correspondence circles but I see no reason for Black
to have sleepless nights over this. With accurate
defence, Black easily refutes the gambit. I keep
wondering why all these strange gambits seem
to have a special attraction for correspondence
players.
9...£k6!
Natural, and best. Black avoids being pinned
by ^.e3 and attacks e5. 9...g6 has been
criticized, and maybe rightly, but the reason isn't
obvious: 10 ±e3 ±c5 11 £ic3 £ie7 12 Scl
(considered White's best) 12...£>b3 (Krantz),
and now:
a) 13 £ixf7? 0-0! 14 Wxb3 ±xe3 +.
b) 13 Wxb3?! ±xe3 14 fxe3 Wxe3+ 15 *hl
Wxg5 16Wxb7Hd8T.
c) 13 £ice4! dxe4 14 Bxc5! (14 £ixe4?!
±xe3 15 &f6+ *d8 16 ±e4 £)d5! was better
for Black in Plato-Johannsson, corr. 1986)
14...^xc5 15 ^xe4 ±c6 (15...Wxb2? 16 £tf6+
*d8 17 ±xc5 £)d5 18 Wf3 Wxe5 19 £ixd7
*xd7 20 Wxf7+ *c8 21 Scl +-) 16 ±xc5
#xb2 17 £tf6+ *f8 18 ±e4 #xe5 19 i.xc6
Wxf6 20 ±xb7 Be8 21 ±c6 Bc8 22 Wd7 Sxc6
23 Wxc6 *g7 24 Wc7 &d5 25 Wxa7 •.
10 Sel (D)

50
French: Advance and Other Lines
10...i.e7!?
Another promising option is 10... JLc5 11 Wf3
£>h6 12 £ic3 £M4!? 13 Wf4 £idf5 14 i.d2 0-0
15 b4 JLe7 + Schmedders-Wilde, Sueder 1996.
Il£ic3i.xg5
ll...a6 12 £ixh7!? *d4 13 £ig5 Sc8 (or
13...£ixe5 14 £ixe6! Wh4 15 h3 i.xe6 16 Sxe5
oo) 14 «M3 Wa7 15 Af4b5 16Scl Ac5 17 Ag3
£\ge7 with roughly equal chances, Poley-Boe,
Tula 2000.
12 i.xg5 a6 13 a3 £ige7 14 Wg4 Wxb2 15
JLxe7 £\xe7
White has no compensation for the
sacrificed material, Masserey-Dgebuadze,
Rotterdam 1998.
B)
9^xd4«xd4(D)
10£ic3
This gambits a second pawn for quick
development. The alternative is 10 We2, after which
White can again follow up by sacrificing the e-
pawn with JLe3 followed by advancing the f-
pawn, or he can move around more slowly with
^hl and f4. Now:
a) 10...f6 has been played rarely but it was
suggested by Keres and is also Watson's
recommendation. 11 ±e3! ? (11 exf6 £>xf6 12 £>c3
i.d6 13 £ib5 We5 + Watson) ll...Wxe5 12 f4
Wb8?! (12..,Wd6 13 Wh5+ <4>e7 looks better,
intending ...£ih6 and ...g6) 13 Wh5+ <4>d8 14
£ic3 £ie7 15 Sacl (15 Sael!?) 15...£>c6 16 f5
d4 17 fxe6 i.xe6 18 Af4 Ad6 19 i.xd6 Wxd6
20 £ie4 We5 21 Wh4 and Black is still not out
of danger, Boschma-Oohmen, corr. 1999.
b) 10...£te7 (I would still prefer this idea).
Then:
bl) 11 <4>hl £ic6 (ll...£>g6 12 f4 ix5 13
£\c3 0-0, Meduna-Jirovsky, Prague 1996,
intending ...f6, is another good idea) 12 f4 £\b4
13 i.b5 0-0-0!? 14 £ic3 a6 15 i.xd7+ Sxd7 16
a3 £ic6 17 i.e3 Wc4 18 Wxc4 dxc4 19 £ia4
Sd3 20 i.b6, Popela-Blaha, Czech Cht 1998,
and now 20...g5! is good for Black.
b2) 11 £>d2 £>c6 12 «M3 Wb6 13 Ae3 Ac5
14 i.xc5 Wxc5 15 Sacl Wb6 ¥ Nikolov-
Dochev, Varna 1995.
b3) H^c3fDjandnow:
b31) n...2L6-10thc3a61lWe2the7.
b32) 1l...£ig6 12 Ae3 Wxe5 13 f4 Wb8 14
f5 JLd6 15 JLb5 JLc6 needs further
investigation but worked well for Black in Lemmers-
Pliester, Dutch Cht 1993.

6 Jid3 AND THE MlLNER-BARRY GAMBIT
51
b33) 1 l...£ic6 12 ±e3 «xe5 13 f4 *d6 14
f5!? (although not convincing, this is a better
try than 14 £ib5 Wb8 15 f5 e5 + Stiirm-Schmid,
corr. 1954-6, when White has less pressure
against Black's centre) 14...«e5! (14...e5?! 15
Sadl ±e7 16 ±c4! is less clear; e.g., 16...d4 17
£>e4 *c7 18 ±xf7+! <4>xf7 19 Wh5+ 4>g8 20
f6! with a strong attack for White, Bateman-
Hellsten, corr. 1987) 15 »f3 (15 »f2 £.d6 16
g3 «f6 +) 15...±d6 16 «h3 exf5 17 Axf5
Axf5 18 Sxf5 *e6 19 £ixd5 0-0 secured a
relatively safe extra pawn for Black in Pinal-
Vilela, Havana 1985.
We return to \0foc3(D):
Now:
Bl: 10...a6 51
B2: 10...«xe5 52
The voluntary retreat 10...Wb6 is also seen
but gives White a freer hand: 11 Wg4! £te7
(Il...h5!?12«g5«d813^ib5«xg5 14i.xg5
f6 * J.Watson-Popel, Lincoln 1969) 12 ±e3
Wd8 13 £ib5 i.xb5 14 ±xb5+ £ic6 15 Sac 1
Sc8 16 Wa4 ± Niermann-Esser, German
Universities Ch 1988.
Bl)
10...a6
Declining the second pawn is a solid choice
but not quite as testing as 10...Wxe5.
H»e2&e7(D)
Black has tried many moves in this position
but the text-move remains the trusted line. An
alternative is ll...Sc8. Black wants to play
...Ac5 without fearing JLe3 but White plays 12
^hl followed by f4 with reasonable
compensation.
1 l...Wh4 is not a bad alternative either, and
has done reasonably well. 12 f4 and now:
a) 12...Ac5+13Ae3(13*hl5ih6ismuch
better for Black) 13...±xe3+ 14 *xe3 £ih6 15
h3 0-0 16 £te2, Smirnov-Leisle, Novokuznetsk
1998, and now, for example, 16...f6 with the
better game for Black.
b) 12...£ih6 13 ±e3 Sc8 14 Sf3 Ac5 15
±xc5 Sxc5 16 Safl «e7 17 Sh3 £>f5 18 »f2
and White had reasonable compensation in the
game Rozentalis-Epishin, USSR Ch
(Leningrad) 1990.
'SS////S _ //////// m. /AU///S. 1 ///Z£///. ' I
12<&hl
Preparing f4 is the most dangerous. 12 Sdl
has been tried as well but does little more than
tease Black's queen a little. 12...£fc6 13 ^.xa6
«xe5 14 «xe5 (14 i.xb7 «xe2 15 £>xe2 Sb8
16 ±xc6 ±xc6 % Bonner-Whiteley, British Ch
(Blackpool) 1971) 14...£ixe5 15 i.xb7 Sa7 16
i.xd5 exd5 17 Sel f6 18 f4 ±c5+ 19 <4>hl d4
20 £ibl (20 fxe5 dxc3 21 exf6+ <4>f7 22 fxg7
<4>xg7 +) 20...0-0 21 fxe5 fxe5 22 £id2 ix6
(22...±b4!?) 23 Sxe5 Sf2! 24 Sxc5 Se7 25
<4>gl Sxg2+ 26 <4>fl Sf7+ 27 <4>el Se7+ 28
<4>fl Sf7+ 29 <4>el d3! 30 Sxc6 Sgl+ 31 £ifl
Sfxfl+ 32 <4>d2 Sdl+ 33 <4>e3 d2 34 ±xd2
Sxal + Sarink-Despotovich, corr. 1979.
12...£ic6
12...*Tb6!? 13 f4 g6 14 ±e3 Wd8 gave White
some play in Smirnov-Temirbaev, Omsk 1996

52
French: Advance and Other Lines
but Black's position is solid and he remains a
pawn up.
13 f4
White can give up a second pawn for a quick
attack by 13 ±e3!? Wxe5 14 f4 !ti6, but after
both 15 f5 We5! and 15 Sadl ±e7 16 f5 We5!,
it seems that Black neutralizes the attack.
13...£ib4 14 Sdl (D)
14 i.bl Wc4 15 ®f3 d4 is clearly not very
attractive for White, Denckens-Harding, corr.
14...±c5!
This allows White to get his pawn back but
Black avoids ending up with a bad bishop
against a knight. There are now a number of
bishop discoveries but Black will simply trade
queens by ...Wf2, with a slightly better
position.
14...£ixd3 15 Sxd3 Wb6 (this is stronger
than the other main continuation, 15...Wc4, on
which White will play 16 b3 Wc7 17 ±b2
followed by f5 with excellent play for the pawn)
16 JLe3 JLc5 (almost invariably played but it is
not clear that it is best, since after an exchange
of the dark-squared bishops Black will have
nothing to challenge a white knight on d4;
16...Wa5 or 16...®d8 might also be considered)
17 ±xc5 ®xc5 (D) and now:
a) 18 f5 is the most direct but maybe not
best. 18...i.c6! (18...0-0-0 19 £>e4!; 18...exf5
19 £ixd5 £b5 20 ®e3 ®xe3 21 £ic7+ <4>e7 22
£\d5+ <£>f8 23 Sxe3 and now Knaak gave
23...g6 24 e6 <*>, but 23...Sd8 24 £>c7 ±c6 25 e6
<4>e7! is good for Black) 19 Wg4 0-0-0 (19...d4
looks even better; it makes sense to upgrade the
c6-bishop even if it gives back the pawn) 20
fxe6 fxe6 21 £ie2 (21 £ia4!? ®a5 22 Scl <4>b8
23 £>c5 d4 * Knaak) 21...<4>b8 22 £id4 with
compensation for White, Pirrot-K.Muller, Bun-
desliga 1989/90.
b) 18 Wd2!? (rather than going for a quick
f5, this seems stronger as White will prevent
Black from playing ...d4) 18...±c6 19Scl ®e7
20 £te2 (White is about to achieve his ideal
set-up with a powerful knight on d4) 20.. .0-0 21
£d4 f6 22 Se3 Sac8 (22...±d7!?), T.Sorensen-
Greger, Copenhagen 1998, and now 23 f5!?
±d7! 24 Seel fxe5 25 Sxe5 Wf6 26 £if3 is
unclear.
15 ±xh7
15 i.xa6 Wf2 16 ®xf2 ±xf2 17 ±e2 0-0 18
±d2 £>c2! 19 Sabl £>e3 20 Sdcl Sfc8 ¥
Organdzhev-Drasko, Skopje 1992.
15...Wf 216 Wxf 2 Axf 217 ±bl 0-0-018 h3
f6!
Black is better, Bellet-Apicella, French Cht
1995.
B2)
10..:g?xe5
The acceptance of the second pawn queries
the correctness of the Milner-Barry Gambit,
provided that Black knows what to do.
11 Sel Wb8! (D)
This is the correct retreat. 1 l...Wd6 12 £\b5!
®b8 (12...Wb6 13 Ae3 Wa5 14 ±d2 Wb6 =)
13 »f3 ±d6 14 £>xd6+ Wxd6 15 Af4 »e7 16
Wg3 ± Messere-Sheikh, corr. 1962.

6 Jid3 AND THE MILNER-BARRY GAMBIT
53
12ftzd5A.d613Wg4
This is the only move that causes Black a few
problems. After 13 Wh5 <£>f8, White must move
his knight away from d5, after which Black can
reply ...£tf6 with a good game.
13...<£f8
The greedy 13..JLxh2+? 14 <£>hl &e5 is
wrong, owing to 15 Af4! +-.
14£d2!?
White should keep his knight on d5 for as
long as possible. Worrying about the h2-pawn
would be a waste of time, and White is now
planning to increase the pressure against Black's
kingside with ic3 or &.M.
14...H5
I am surprised how little attention has been
paid to the greedy 14...£xh2+ 15 <£>hl £d6
since it isn't that obvious how White can best
strengthen his attack. However, one possible
line is 16 i.c3 f6 17 Sadl h5 18 Wh3 #c8 19
&e3 &e7 20 £e4! e5 21 Wh4 £c5 22 f4 with a
strong attack for White. Another greedy attempt
is 14...f5 but this also appears very dangerous
for Black after 15 £xf5! (15 Wf3!? exd5 16
Axf5 is also interesting) 15...exf5 16 Wd4, as
in Sziebert-Stiemer, Balatonbereny 1994.
15#h3Ac6fD;
Not the only move but a very logical one as
White's strong knight is forced away. Black
should avoid 15...exd5?! 16 #xd7 £xh2+ 17
<£>hl &f6 18 Wf5 £d6 19 Sacl, when White
had very good compensation in Borg-P.Nikolic,
Kavala Z 1985, but 15...£ih6 and even just
15...£if6!? are also playable.
16 <&e3
After 16 &b4 a few games have concluded
with a repetition by 16...£d7 17 £id5 £c6 18
£lb4, but Black came out on top in Burnett-
J.Watson, Los Angeles 1982 after 16..JLe8 17
£c4 «M6 18 £>d3 £>g4 19 g3 b5 20 £b3 £c6!
+.
16...&f617<&c4.£.c7
Kuznetsov's 17...£ld5 is also good.
18 £c3 &g4! 19 g3 b5 20 &a5 if3
Black is better - J.Watson.

4 5...J.d7
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 £k6 5 £>f3 ±d7
(D)
5..Adl is the main alternative to 5...#b6. It
is more flexible but also in some lines less
forcing, and thus White can usually complete his
development without the usual worries about
how to defend d4. White's main choice is
whether to aim for a quick queenside expansion
with a3 and b4 anyway, which also provides
White with a way to defend d4 with his bishop
from b2, or simply to get on with development
(6 JLe2). In the latter case, White has the option
of lending support to the defence of the d-pawn
with £ia3-c2.
The choice is between:
A: 6dxc5 55
B: 6a3 56
C: 6±e2 63
Some minor alternatives:
a) 6 ±e3 Wb6 7 Wd2 - Kupreichik's 5 k.e3
Wb6 6 Wd2 ±d7 7 fof3 (Chapter 6).
b) 6 .&d3 invites Black into a Milner-Barry
Gambit after 6...Wb6 but as explained in
Chapter 3, this move-order allows White to avoid
sacrificing the pawn with 7 0-0, and instead
choose 7 dxc5. Hence Black has tried other
moves, the most popular being 6...Sc8 (D)
(6...cxd4 7 cxd4 £>b4 8 ±e2 only results in
Black's knight being kicked back again with a
later a3 but is certainly a viable option as White
might also feel tempted to move his bishop
back to d3, so the tempi wasted cancel each
other out). Now:
bl) 7 a3 £>h6 8 b4 cxb4 9 cxb4 £>f5 -6a3
Sc8 7 b4 cxd4 8 cxd4 *hge7 9 ±d3 Qsf5.
b2) 7 0-0 cxd4 8 cxd4 £>b4 9 £>c3 £>xd3 10
Wxd3 £ie7 is fine for Black due to his bishop-
pair, although White's advantage in
development shouldn't be underestimated.
b3) 7 dxc5 ±xc5 8 0-0 £>ge7 9 b4 ±b6 10
Sel (10 b5 £>a5 - 6 dxc5 ±xc5 7 b4 kb6 8 b5
tha5 9 ±d3 *he7 10 0-0 ScS) 10...£>g6 11
±g5 Wc7 12 £>a3 a6 13 2c 1 £>cxe5 14 £>xe5
£>xe5 15 ±bl £>c4 16 £>xc4 Wxc4 17 Wh5
JLd8! and any hope White might have of
getting compensation for the sacrificed pawn has
been firmly repelled, Kislinsky-Firman,
Kharkov 2001.
c) 6 g3 Wb6 7 h4 (7 ±h3 might be better but
looks odd with the bishop in front of the pawn;
if Black wants he can tease a bit with ...Wa6 but
there should be other ways too) 7...f6! 8 Jih3
cxd4 9 cxd4 fxe5 (9...^.b4+ should probably be

5..±d7
55
met by 10 ^f 1 and maybe even sacrificing the
f-pawn in the event of 10...fxe5 11 dxe5 JLc5,
since 10 £\c3 fxe5 11 dxe5 d4 is just bad for
White) 10 dxe5 (the general advice in similar
situations is to recapture with the knight but
here 10£ixe5ismetby \0..Mx&4) 10...±c5 11
0-0 £ige7 12^c3 £id4 13 £ixd4 ±xd4 14 <4>g2
0-0 15 f4 £>f5 + Szittar-Galyas, Hungarian Cht
1994/5.
A)
6 dxc5 (D)
I
▲ WW A WW A WW/ ▲
This seems to help Black develop, but White
intends to follow up by advancing his b-pawn,
thereby making more room for his minor pieces
in the centre by forcing Black's bishop and
knight to move away.
6...±xc5 7 b4
White's main idea is to attack Black's c6-
knight with the b-pawn, which will give White
better central control. However, the drawback
to this is that his queenside becomes very loose.
The advance of the b-pawn can also be held
in reserve by 7 JLd3, thus not committing the
queenside so soon. Then:
a) 7...b5!?-Sveshnikov.
b) 7...£>ge7 8 0-0 (maybe better is 8 b4 ±b6
9 b5 £ia5 -7b4 kb6 8 b5 £ha5 9 k.d3 *he7)
8...£>g6 9 lei Wcl 10 We2 0-0 (10...f6!?) 11
b4 ±b6 12 a4 a5 13 b5 £ice7 14 Sa2!?
intending Ec2 and c4 with an edge for White, Cutillas
Ripoll-John, corr. 1990.
c) 7...f6!?8b4(T>,)andnow:
cl) 8...±b6!? 9 b5 £ixe5 10 £ixe5 fxe5 11
!Ti5+ <4>f8 12 ±a3+ £>e7 13 Wxe5 <4>g8 14
£>d2 £>g6 15 Wg3 Wf6 16 0-0 Wxc3 17 £>b3
We5 (17...£>f4 18 Wxf4 Wxd3 +) 18 Wg4 £>f4
19 Sael!? Axf2+ 20 Sxf2 Wxel+ 21 .fi.fl,
Vysochin-Thompson, Groningen 1996, and
now 21...£>d3 22 Se2 Wbl! leaves White with
insufficient compensation.
c2) 8...±e7 9 b5 £ixe5 10 £ixe5 fxe5 11
Wh5+ <4>f8 12 Wxe5 and then:
c21) 12...±f6 13 Wg3! e5 14 0-0 £>e7 15
.fi.g5 is better for White according to Lane.
c22) 12...£if6 13 0-0 Wb8 14 f4 <4>f7 15
±e3 (15 We2 ±d6 - 72... «*8 13 ft! ^hf6 14
We2 ±d6 15 0-0 */7) 15...£d6 16 Wg5 Sc8
17 ±d4 ±c5 18 £>d2 ± Roca-Kasimdzhanov,
Erevan OL 1996.
c23) 12...Wb8 13 f4! £>f6 14 We2 ±d6 15
0-0 <4>f7 16 £id2! Sf8 (16...i.xf4 17 £>f3 with
compensation) 17 £rf3 4^8 18 £>g5 ±c5+ 19
<4>hl Wd6 20 c4 ± S.Pedersen-Cu.Hansen,
Odense (training game) 1995.
7...±b6
7...±e7 8 b5 £ia5 9 ±d3 Wcl 10 We2 is
similar to the main line but the bishop should be
better placed on b6 where it looks towards the
centre and White's kingside.
8b5
White need not play this immediately but
can also simply play 8 JLd3. However, this may
give Black a chance to prevent b5 by playing
...a6, or 8...f6!? - 7 ±d3f6!? 8 b4 kb6l?.
8...£la5 9 ±d3 (D)
The drawback of having the bishop on b6
instead of on e7 is that in some lines White's

56
French: Advance and Other Lines
m k i
._ iiilil
ftp HH A iH if
"" bi
mm i
ft • IP H ft«
Ifl few
dark-squared bishop is going to appear on a3
with strong effect.
9...£>c4!?
This is a safe choice for Black. He attempts
to prevent White's bishop from getting to a3,
and attacks the b5-pawn. 9...&e7 10 0-0 Sc8 11
a4 £>g6 12 Sel f6 13 Sa2!? 0-0 14 ±xg6 hxg6
15 Wd3 <A>f7 16 ±e3 gave White an edge in
Sveshnikov-Naumkin, Moscow 1989.
10 We2 Wc7 11 ±xc4!?
An attempt to improve upon 11 a4 a6 12
bxa6 Sxa6 13 0-0 £>e7 (13...Sxa4 14 Sxa4
JLxa4 15 £te3 with compensation) 14 £&3 Sxa4
15 £>xc4 dxc4 (15...2xal 16 £>d6+ <A>f8 17
£>g5 ±e8 oo) 16 Sxa4 ±xa4 17 ±xc4 0-0 18
±d3 £>g6! (V2-V2 Kharlov-Dreev, USSR Ch
(Moscow) 1991) 19 h4 ±c6 20 h5 &xf3 21
gxf3 Wxe5 22 hxg6 Wg3+ 23 <4>hl Wh3+ 24
<A>gl V2-V2 Kharlov-Svidler, Novosibirsk 1995.
All this had even been analysed before by Dreev.
Il...dxc4
11...WXC4 12 Wxc4 dxc4 13 £>a3 Sc8 14
£M2 is given as slightly better for White by
Zakharevich.
12 £>bd2 ±xb5 13 £>e4 ±c5!
13...0-0-0? 14 £>d6+ Sxd6 15 exd6 Wxd6 16
£ie5 ± Zakharevich.
14 a4 ±c6! 15 £>xc5 ±xf3 16 Wxf3 Wxc5
17 0-0 Wd5!
The chances are roughly equal, Kharlov-
Zakharevich, Russian Ch (Elista) 1996.
B)
6 a3 (D)
This move is White's most popular reply
against 5...Wb6. Against 5.. JLd7, 6 a3 also has
its place, and the threat of queenside expansion
is as annoying here as it is with the queen on
b6. However, the flexible .. JLd7 gives Black
slightly more ways to seek counterplay than he
has with the queen already on b6. We shall
examine four replies:
Bl: 6...Sc8 56
B2: 6...£>ge7 58
B3: 6...C4 59
B4: 6...f6!? 61
6...a5 needs a mention too. After 7 JLd3,
7...1rb6 transposes to 5...Wb6 6 a3 a5 7 k.d3
kd7, while 7...a4?! 8 dxc5! ±xc5 9 0-0 is good
for White.
Bl)
6...Sc8 7 b4
A very complicated position arises from 7
±d3 cxd4 8 cxd4 Wb6 9 Ac2 (D) (9 0-0 £>xd4
10 £>xd4 Wxd4 11 £>c3 is feasible a la Milner-
Barry Gambit but Black's ...Sc8 looks more
useful than White's a3) and now:
a) 9...£>xd4?! 10 £>xd4 ±c5 11 £>e2 (11
£>f3 ±xf2+ 12 <4>e2 should be very good for
White) ll...±xf2+ 12 <4>fl f6 13 ±a4 fxe5 14
±xd7+ <4>xd7 15 g3 Bf8 16 Wa4+ <A>c8 17 <4>g2
with an unclear position, Afek-Krays, Rishon
leZion 1993.
b) 9...g5! 10 h3 (10 £>xg5? £>xd4 11 ±d3
£>b3 -+; 10 ±xg5? Wxb2) 10...£>xd4!? (the
same idea from above but the inclusion of h3

5...kd7
57
iAB
I
V/,^W//,
and ...g5 clearly favours Black) 11 £\xd4 JLc5
and then:
bl) 12 £>f3? Axf2+ 13 <4>e2 ±g3 (the main
point of provoking h3) 14 Sfl? 2xc2+ 15 Wxc2
#a6+! -+.
b2) 12±e3?lrxb2 13i.b3(13 0-0i.b6+)
13...«fxal 14 0-0 £ie7 15 Sel £ic6! 16 £ixe6!
(16£>c2Wb2 17Axc5Wxb3-+; 16^b5±xe3
17 £>d6+ <4>e7 18 £>xc8+ Sxc8 19 Sxe3 £>xe5
-+) 16..JLxe3 (according toFta£nik, 16...^.xe6
17 JLxc5 b6 is safer, with a clear advantage to
Black) 17 £ig7+ <4>f8! 18 Wxd5 £ixe5 19 Sxe3
2cl+ 20 9feh2 Sxbl, Hodgson-Yusupov, Gro-
ningen 1994, and now 21 Sxe5 is White's best
practical chance, although Yusupov gives a
convincing refutation in 21...2hl+ 22 <A>g3 2xh3+!
23 gxh3 Wgl+ 24 Wg2 (24 <A>f3 ±c6 is also
winning for Black) 24...1rxg2+ 25 *xg2 <A>xg7
26 2xg5+ 9fef6, when Black should win the
endgame.
b3) 12 £ie2 Axf2+ 13 *fl f6! (this is the
point of Black's attacking idea: the f-file is
opened against the white king) 14 JLa4 fxe5
and now:
b31) 15 ±xd7+? (Black's attack is only
accelerated after this mistake) 15...9frxd7 16 Wa4+
Sc6 17 g3 £>e7 18 *g2 £>f5 gave Black a strong
attack in Sveshnikov-Chernin, USSR Ch (Riga)
1985.
b32) White should instead try 15 £>bc3,
when 15...£te7? is met by 16 ±xd7+ <A>xd7 17
<£>a4 +-. However, Black seems in very good
shape after 15...£c5! 16 ±xg5 £ie7 17 b4 0-0+
18<A>eli.f2+19<A>d2±c6!.
7...cxd4 8 cxd4 £>ge7 (D)
This is a logical way to develop but 8...Wb6,
transposing to 5..Mb6 6 a3 ±d7 7 b4 cxd4 8
cxd4 2c5, is probably better.
W
Jill
9 ±d3 (D)
Or:
a) 9 ±b2 £>f5 and now 10 ±d3 transposes
to the main line, and is possibly a more accurate
way to reach it. White can also try 10 h4!? JLe7
11 h5.
b) 9 £>bd2 Wb6 10 £>b3 £>b8! 11 Ad3 (11
£>c5 ±b5 =) 1 l...±a4 followed by ...£>bc6 and
...a5 was fine for Black in Labunsky-Kiriakov,
Moscow 1996.
^kWkMkWtk
//"'i'-' * %%% ^js^
9...£fcf5
9...£>g6 10 h4 kzl 11 h5 ^h4 12 £ixh4 ±xh4
13 ±e3 ±e7 14 Wg4 ± Kucera-Felix, Karvina
1989, but 9...b5!?, as in Sarwat-Arizmendi,
Tanta 1998, intending to play on the queenside
with ...Wb6 and ...a5, is maybe playable.

58
French: Advance and Other Lines
10 ±b2 (D)
10...i.e7
10...#b6 looks better. Then 11 ±xf5 exf5
12 £\c3 JLe6 13 0-0 JLe7 was only
microscopically better for White in Parmensini-Panno,
Mar del Plata 1996.
110-0g5!?
More active than 11...0-0 12 £>c3 g6 13 £>e2
£>g7 14 Wd2 f5 15 exf6 i.xf6 16 £>e5 Ae8 17
f4, when White looked firmly in control in
Motylev-Cao, Ohrid Ech 2001.
12 g4
12 JLxf5 exf5 13 £\c3 JLe6 is a common type
of position, which is probably roughly level.
12...£>h6 13 h3 0-0 14 £>c3 f5 15 exf6 i.xf6
16 £>e2 i.g717 £>g3 Wffi18 £>e5! i.e819 Scl
± Kalod-Kuporosov, Pardubice 1998.
B2)
6...£>ge7 (D)
This is a little more flexible than 6...Sc8
(Line Bl), although the two can transpose if
Black plays ...2c8 later.
7b4
White has also tried two other moves:
a) 7 dxc5 is rarely seen, and bearing in mind
that 5...£hge7 6 dxc5 isn't regarded as very
favourable for White, it is difficult to believe that
the inclusion of a3 and ...JLd7 should make
things look better for White.
b) 7 JLd3 cxd4 8 cxd4 and now Black should
play ...£tf5 immediately, or after 8...1rb6 9 ix2.
White is going to take on f5 anyway, so it is a
W
f Wf ^ JlL M.
^■kWkM'kMk
I
question of whether Black would like his queen
on d8 or b6. On b6, the queen is seemingly more
active but may also be exposed to an attack after
£\c3-a4, whereas from d8 the queen might assist
in grabbing space on the kingside with ...g5. The
most popular line is 8...ffo6 9 ix2 £tf5 10
JLxf5 exf5 11 £\c3 JLe6. A few games have
continued 12 b4?! Sc8 13 Wd3? £ixb4!, winning a
vital pawn, since 14 axb4 is met by 14...Sxc3!
15 Wxc3 JLxb4. This serves as a warning that
White should be careful when he expands, and
thereby also exposes, his queenside. In Masse-
rey-Dizdar, Geneva 1999 White continued more
safely: 12 £>e2 h6 13 h4 g6 14 £>f4 i.e7 15 Sh3
4^7!? with roughly equal play.
7...cxd4 8 cxd4 £>f5 (D)
W
^-fcwi —saw*
irJmkp^hm^
mm mm w§ a m
l§ H o ii is
9±b2
White has some tactical problems after 9
£>c3? flc8 10 i.b2 £ih4! and, in Illescas-Speel-
man, Linares 1992, got into severe difficulties

5...±d7
59
following 11 £\xh4?! (11 £\a2 is more prudent,
not allowing Black's queen to join the attack
immediately) ll...Wxh4 12 £>a2 (not 12 £>e2?
^xb4! -+) 12...We4+! 13 We2 Wg6! 14 Scl
±e7 15 Wf3 0-0 16 ±d3 #h6 17 0-0 f6!, with a
clear advantage to Black.
9 JLe3 is also possible but Black equalizes
comfortably with 9...£\xe3 10 fxe3 g6!; e.g., 11
±d3 ±h6 12 We2 f6! 13 exf6 Wxf6 14 0-0 0-0
15 £>c3 We7 16 Sael Sac8 = Kalod-Zuger,
Portoroz 1998.
9...b5!? (D)
With this Black initiates play on the queen-
side, the tactical point being that 10 Jtxb5 is
met by 10...^xe5. More on this in the next
note. Black can also play:
a) 9...*b6 - 5..Mb6 6 a3 £hh6 7 b4 cxd4 8
cxd4 Zhf5 9 kb2 k.d7.
b) 9...Sc8 10 ±d3 - 6..Mc8 7 b4 cxd4 8
cxd4 Zhge7 9 k.d3 £>/5 10 kb2.
c) 9...&e7 10 ±d3 0-0 11 0-0 f6 12 Axf5
exf5 13£fc3£e8!?(13...£e6 14Bel ±) 14Wd3
f4 15 e6 Jth5 16 Ifel ± RNikolic-Christiansen,
Portoroz/Ljubljana 1985.
10 £>bd2
Black has equalized easily against this but
other moves have not promised White anything
either:
a) 10 ±xb5 £ixe5 11 £ixe5 ±xb5 12 £>c3
JLd7 was approximately equal in Kalezic-
Cabrilo, Yugoslav Cht (Herceg Novi) 2001.
b) 10 ^c3 and then:
bl) 10...a5!? 11 ±xb5 (11 £ixb5 axb4 12 a4
^a7! =) 1 l...axb4 12 axb4 ±xb4 13 0-0 0-0 14
±d3 Sxal 15 Axal Wa5 = Imandi-Yulianto,
Asian jr Ch (Vung Tau City) 1999.
b2) 10...a6 11 ±d3 g6 12 0-0 h5 13 £>e2
±e7 14 ±c3 Wb6 15 £>f4 £>g7 16 h4 a5 17
We2 axb4 18 axb4 Sb8 19 ffifbl 0-0 = Grosar-
L.B.Hansen, Bled/Rogaska Slatina 1991.
10...a5 11iLxb5axb412a4
12 axb4 £>xe5 13 ±xd7+ £>xd7 ? Psakhis.
12...±e7 13 0-0 0-014 h3 £>a515 JLd3 2c8
16 We2 Wb6 17 g4 £>h6 18 ±bl Sc7!
Black intends ...Sfc8 with counterplay, Mov-
sesian-M.Gurevich, New York Open 1998.
B3)
6...C4 (D)
This is another common idea in the Advance
Variation: Black closes the queenside, hopes
his kingside survives and argues that there will
be counterchances if he can exploit the
weakening of White's queenside. However, I tend to
agree with Kosten that this idea is more
favourable for Black in the 5...#b6 line, because
White's dark-squared bishop is then tied to the
defence of b2. Here White gets the chance to
develop the bishop, which he should take. If he
doesn't, Black's queen will be better placed
back on d8 compared with the 5...Wb6 6 a3 c4
lines.
7±f4
This seems best. Other moves:
a) 7 b4 cxb3 8 Wxb3 is played surprisingly
frequently. White avoids having his queenside
fixed but at the costly price of clearly weakening

60
FRENCH: ADVANCE AND OTHER LINES
his pawn-structure. I believe Black is better after
8...£>a5 9 «c2 £>e7 10 Ad3 1^6!?, intending
..JLb5, as in Vennix-Hoekstra, corr. 1987.
b) 7 Ae3 is frequently played by Afek.
Compared with the main line (7 Af4) the
bishop does not obstruct a future f-pawn
advance. White also avoids ideas where Black
gains some time on attacking White's bishop
and knight by ...g5-g4. Nevertheless, I don't
believe it to be better than 7 Af4. Both 7...f6 and
7...trb6 8 #c2 f6, which to my knowledge
haven't been played, are worth testing. Instead
Afek-Giddins, Amsterdam 2001 went 7...^a5
8 £>bd2 ±c6 9 h4 Wdl 10 h5 h6 11 £>h4 Aa4
12 ©bl 0-0-0 13 f4, maybe with an edge for
White.
c) 7 h4 (D) and now:
cl) l..Mcl 8 h5 0-0-0 was a success for
Black in Ivanovic-Levitt, Saint John 1988 but
only after White's inaccurate 9 £>h4? f6! 10 f4
g5! 11 £>f3g412£>h2f5+. 9b3!?isamore
serious test of Black's set-up.
c2) 7...£>a5 8 £>bd2 Ac6 9 h5 #d7 10 ^h4
0-0-0 11 g3 f6 (ll...Aa4!? 12 «e2 Ac2 is a
common idea; following 13 £>xc4! £>b3 14
«xc2 £>xal 15 £>b6+! *c7!? 16 fdl <4>xb6 17
Ae3 White will have some compensation after
he collects the knight) 12 f4 £>h6 13 #e2 f5 14
Sbl tte8! 15 £>g2 g6! 16 £>e3 Ae7 <*> Haba-
Ruzele, Lubniewice 1994.
d) 7 g3 £>a5 8 £>bd2 ±c6 9 Ag2 Wd7 10
®e2! (if 10 0-0, Black favourably improves his
bishop with 10...Aa4 11 «e2 Ac2!) 10...Aa4
11 £>fl £>b3 12 Sbl £>xcl 13 Bxcl 0-0-0 14
£>e3 4b8 15 0-0 £>e7 16 £>g5 £>c8 17 f4 ±
Haba-Psakhis, Pardubice 2002.
We now return to 7 MA (D):
I i'ti4I
7...£>a5
With slightly the better structure on the
queenside, Black sends his minor pieces in that
direction in an attempt to infiltrate on the light
squares. We have seen the same idea in a
number of other lines above. Two other options are
worth mentioning, the first being very similar:
a) 7...£>ge7 8 h4 £>c8 9 £>bd2 £>a5 10 h5 h6
11 g4!? Ml 12 Sgl £>c6 13 g5 hxg5 14 £>xg5
± Markovic-Pavlovic, Yugoslav Cht (Herceg
Novi)2001.
b) 7...trb6 8 «c2 f6!? 9 £>bd2 (9 h4!? is
maybe better) 9...g5! 10 Ag3 g4 11 £>h4 fxe5
12 dxe5 and now:
bl) 12...0-0-0 13 h3 k.%114hxg4(14 Axc4
Axe5 15 Ml Axg3 16 fxg3 £>f6 ?) 14...£>xe5
15 ^hf3 £>xg4 16 Axc4! £>e7 17 M3 with the
better chances for White, Hodgson-Shabalov,
Winnipeg 1997.
b2) 12...£>h6! 13 Ml (13 h3 k.gl 14 hxg4
Axe5 15 Axe5 £>xe5 16 g5 £>hg4 17 £>df3
0-0-0 18 £>xe5 £>xe5 19 Ml Sdf8 20 £>f3
£>g4+; 13 0-0-0!?) 13...Ag7 14h3?(14 Axc4!
has the idea that 14...dxc4 15 £>xc4 «a6 16
£>d6+ *d8 17fd2 <4>c7 18 MA gives White a
strong attack for the piece, but according to
Gulko Black is better if he declines the sacrifice
by 14...£>xe5! 15 Ae20-0!) 14...Axe5 15 0-0-0
Axg3 16 fxg3 0-0-0 + Sveshnikov-Gulko, Vol-
godonsk 1981.
8£>bd2±c6(T>)

Black prepares to play on the light squares
on the queenside with ..."#d7, which besides
.. JLa4 prepares ...0-0-0.
9h4
This is White's best plan. The idea is to
induce weaknesses on Black's kingside, either by
advancing the h-pawn all the way to h6, or if
Black replies ...h6 then playing for f4-f5. The
latter plan is advantageously carried out with a
white pawn on h5, because it will then be
impossible for Black to set up a light-squared
blockade with ...g6 and ...h5. Arencibia-Psa-
khis, Andorra 1995 saw White playing more
passively, and Black equalizing comfortably: 9
Ae2Wd7 100-0 Aa4 11 ©el £>e7 12Sel £>g6
13 Ag3 Ae7 14 Sbl h5 15 h3 h4 16 ±h2 0-0-0
17£>flf5! 18exf6gxf6=.
9..Mdl
To prevent White's plan, Black could also
play 9...h5 but Black often elects to refrain from
this in these types of positions. The reason is
that his structure is then largely fixed on the
light squares and White has good prospects of
slowly preparing a breakthrough on the king-
side; e.g., 10 g3 Wd7 11 Ah3 Aa4 12 «cl £>e7
13 0-0 0-0-0 14 Sel £>ec6 15 £>g5 b5 16 Sbl!
4>b7 17 £>fl £>b3 18 «e3 Bc8 19 «e2 a5 20
£te3 with a slightly better game for White,
Degraeve-Le Bideau, Issy les Moulineaux 2000.
10 h5 Aa4 (D)
ll«cl
The queen later goes to bl anyway, so 11
Wbl might be worth considering. From there it
is also more influential on the kingside.
11...H6 12 g4
61
This is very committal, and it is not clear that
White wants to play this at all. 12 Ae2 is more
flexible; White prepares 0-0, and maybe £Mi4,
followed by moving the bishop away and f4-f5.
12...£to7 13 Ae3 £>c8 14 £>h4 £>b6 15 f4
Ae7 16 ttbl ^b3 17 £>xb3 cxb3 18 f5 0-0-0
Afek-McDonald, Oakham 2000. Black is
doing well: he has a positionally sound game
and intends to exchange the light-squared
bishops. If his kingside doesn't crack, he will
always have good chances in the endgame due to
White's bad dark-squared bishop.
B4)
6...f6!? (D)
This fashionable and direct approach has
also gained popularity because of John
Watson's advocacy of it.
7i.d3

62
FRENCH: ADVANCE AND OTHER LINES
With this White seizes the opportunity to
develop the bishop actively on d3. 7 Ae2 is
obviously worse, since in comparison with 6 JsLe2f6
White has wasted a tempo on a3. Other moves
include:
a) 7 exf6 gxf6 8 £>h4 #e7 9 #h5+ Wf7 has
been given as unclear by Dolmatov and
certainly doesn't look bad for Black.
b) 7 &f4?! #b6! 8 b4 cxd4 9 cxd4 g5! 10
Ae3 g4 11 £>fd2 fxe5 12 ^b3 (12 dxe5 #c7 13
#xg4 #xe5 is also much better for Black)
12...exd4 13 £>xd4 Ah6! 14 #xg4 0-0-0 15
£>c2 Axe3 16 fxe3 e5 + Gramer-Djurhuus,
Gausdal 1991.
c) 7 b4 fxe5 8 b5 (8 £>xe5 £>xe5 9 dxe5 #c7
¥; 8 dxc5?! e4 9 ^d4 £>f6 ? Popchev-Dolma-
tov, Polanica Zdroj 1987) 8...£>xd4! 9 £>xe5 (9
cxd4 exd4 10 &f4 ^f6 11 £ibd2 Ae7 12 g3 0-0
¥ Buchal-Luther, Groningen 1998) 9...£>f5 10
#h5+ g6 11 £>xg6 £if6 12 #h3 Sg8 13 £>xf8
&xfS 14 Ad3 £>e4! with counterplay, Movses-
ian-Luther, Cappelle la Grande 1998.
We now return to the position after 7 Ad3
(D):
7...#c7
The direct attack against e5 is logical and is
the basic idea behind 6...f6!?. There are two
other ideas for Black, which both aim to fix a
white pawn on e5 and thereafter attack it:
a) 7...cxd4 8 cxd4 £tfi6 was initially strongly
supported by Watson and Knaak. Knaak later
changed his mind, and admitted he was a little
overoptimistic regarding Black's idea. The
critical variation runs 9 0-0 £tf7 10 Sel! (White
shouldn't take on f6) 10..JLe7 11 £>c3 0-0 12
#c2 and now 12...h6 13 #dl fxe5 14 dxe5
Ae8 15 Abl £>h8 16 £>e2 Sxf3 17 gxf3 Ag6
18 £>g3 Axbl 19 Sxbl £>g6 20 f4 £>h4
offered Black some compensation in Movsesian-
Luther, Groningen 1998. Luther was prepared
to play this line again, so either he believes
Black's compensation to be adequate, or he has
an earlier improvement. Maybe 12...f5 isn't so
bad; in Hitzgerova-Leer Salvesen, Stockholm
2002 Black had fine counterplay after 13 *hl
Sc8 14 #dl Ae8 15 £>e2 £>a5.
b) 7...fxe5 8 £>xe5 £>xe5 9 dxe5 g6 (Watson
suggests 9...#h4!?) also looks like a tough
defence to crack. Afek-Hebert, Paris 1995
continued 10 0-0 #c7 11 f4 c4 12 Ac2 Ac5+ 13 *hl
£>e7 14 ^d2 0-0-0 15 b4 cxb3 16 £ixb3 Ab5
with an unclear position.
8 0-0 (D)
8...c4
It is probably a good idea for Black to play
this before White plays Sel and gains the
option of retreating the bishop to f 1. Black might
also wait with ...c4 altogether, and 8...0-0-0
worked well for Black in Grishchuk-Short,
Reykjavik 2000: 9 #e2 h6 10 b4 c4 11 Ac2 f5
12 £>h4 Ae8 13 f4 Ae7 and Black had solved
his opening problems so well that Grishchuk
went for 14 £>xf5?! exf5 15 Axf5+ *b8 16
#g4, but Short defended well with 16...g5! 17
fxg5 hxg5 18 ±xg5 Ah5 19 #g3 ±xg5 20
#xg5 £}ge7 and went on to win. 9 Bel looks
like a more natural way to develop. It is not
clear then how Black reacts, and after 9...c4 10

5...&d7
63
jLfl, White's bishop is probably better placed
than in the main line.
Note also that it is too late for the idea of
exchanging on e5, because, as pointed out by
Anand, 8...fxe5 is now met by 9 £>xe5 £>xe5 10
Af4 i.d6 11 dxe5 ±.
9 Ac2 0-0-0 10 Sel (D)
An important position. The stage is now set
for attacks on opposite sides but with an
interesting central struggle to complicate matters.
10...fxe5
This capture is the most common but other
moves are also seen; for example, 10...^ge7,
when 11 b3 cxb3 12 i.xb3 £>g6 13 exf6 gxf6
14 c4, as in Shilov-Kiseleva, Legnica 1995,
looks critical. Black should now play 14...dxc4
15 JLxc4 e5! with a complicated position.
11 £>xe5
Preferable to the other recapture 11 dxe5
±c5 12 £>bd2 ^h6 13 b4 i.b6 14 h3 £>f7 15
#e2 Sdf8 16 <4>hl g5 17 a4 a5! 18 b5 £>e7 and
Black was slightly better in Alekseev-S.Ivanov,
St Petersburg 1995.
Il...£>xe5 12 dxe5 Ac5 13 Ae3 i.xe3 14
lxe3 £>e7 15 £>d2 £>c6
15...2df8? 16 #g4! Hf7 17 £>f3 was better
for White in Alekseev-R.Bagirov, St
Petersburg 2000, and is probably a significant move-
order error.
16£if3
16 #g4 looks more critical; e.g., 16...^xe5
17 #xg7 £ig6 18 Sael <4>b8 19 £>f3. In Haba-
Potkin, Pardubice 2002 Black played the safer
16...2dg8 and after 17 £M3 a draw was agreed.
Obviously, this says little, and there is plenty of
play left.
16...Sdf8
The chances are roughly equal, Haba-Uly-
bin, Koszalin 1999.
C)
6 Ae2 (D)
IB B*B*fl
Now:
CI: 6...#b6 63
C2: 6...f6!? 65
C3: 6...Sc8 70
C4: 6...£>ge7 71
Two minor alternatives:
a) 6...£tfi6?! is reminiscent of various ...£tfi6
lines but right here it is criticized by Bareev. 7
i.xh6 gxh6 8 0-0 #b6 9 #d2 Ag7 10 £>a3 0-0
11 £>c2 cxd4 12 cxd4 f6 13 exf6 Sxf6 14 b4
Saf8 15 b5 £>e7 16 £>e5 Ae8 17 g3 led to the
better position for White in Topalov-Bareev,
Novgorod 1997.
b) 6...f5 is a solid type of move but
determining the central structure like this often
operates in White's favour: 7 0-0 cxd4 8 cxd4 £ige7
9 b3 £>c8 10 Aa3 i.xa3 11 £>xa3 0-0 12 #d2
#e7 13 £>c2 Ae8 14 b4 a5 15 b5 £id8 16 Hfcl
was better for White in Illescas-Short, Linares
1995.
CI)
6...#b6 (D)

64
French: Advance and Other Lines
w
I
•111
A if IP & if ft iH
i&hvb
W%, Wa
w
ft Iftf » & Mf ft Ifti
7 0-0
7 dxc5 ^.xc5 8 0-0 is a serious alternative:
8...a5 (8...#c7!?) 9 c4! £>ge7 (9...d4 favours
White after 10 £>bd2 £>ge7 11 i.d3 £>g6 12
Sel) 10 £ic3 £>d4 (again, 10...d4!? would give
White an advantage after 11 &a4 #a7 12 i.d3
£>g6 13 Sel) 11 cxd5 £>xd5 12 £>g5!? h6 (it is
not clear that this is necessary; 12...^.c6 or
12...Sd8 could also be considered) 13 £lge4
£>xe2+ 14 #xe2 £>xc3 15 bxc3 i.c6 16 £>d6+
i.xd6 17 exd6 #c5 18 Sdl favoured White in
Rozentalis-Stanec, Pula Echt 1997.
7...cxd4
Or:
a) 7...f6-6...f6 7 0-0Wb6.
b) 7...£>ge7 8 dxc5 #xc5 (8...#c7 9 b4!?
£>xe5 10 £>xe5 #xe5 11 Sel #c7 12 i.b2 £>g6
13 c4 dxc4 14 £>d2 ±e7 15 £>xc4 was better for
White in Sveshnikov-Kochiev, Le Havre 1977) 9
Ae3 #a5 10^bd2 #c7 (Blatny gives 10...£>g6
11 c4! oo as a better line for Black) 11 JLf4 £>g6
12 ^.g3 £kxe5 (this is risky, and it is possible
that it is better to take with the other knight in
order to have a knight blocking the c-file) 13 c4
i.d6 14cxd5 exd5 15 Sel #b6 16 £>xe5 £>xe5
17 £>f3 £>xf3+ 18 Axf3 ±xg3 19 Sel+! ±e6
20 #a4+ <4>f8 21 hxg3, Blatny-Raicevic, Bad
Worishofen 1988. White is clearly better since
Black cannot coordinate.
c) 7...Sc8 8 dxc5 ±xc5 9 b4 Af8 10 ±d3
£>ge7 11 i.f4 £>g6 12 Ag3 i.e7 13 h4 f6 14
exf6 i.xf6 15 b5 £>a5 16 h5 £>e7 17 a4 0-0,
Castro-Korchnoi, Linares 1979, and now 18
£le5 is better for White.
d) 7...£>h6 (D) and here:
dl) 8 dxc5 ±xc5 9 b4 ±e7 10 Af4 (10
±xh6 gxh6 11 #d2 #c7 12 #xh6 £>xe5 13
#g7 £lg6 was good for Black in Saint Amand-
Rozentalis, Quebec 2001) 10...f6 11 exf6 i.xf6
12 b5 £le7 13 a4, and in this fairly unclear
position a draw was agreed in Timman-Piket,
Amsterdam 1994.
d2) 8 £>a3 cxd4 9 cxd4 i.xa3 (9...£>f5 10
£>c2 - 6...£hge7 7 £ha3 cxd4 8 cxd4 fof5 9 *hc2
Wb610 0-0) 10 ±xh6 ±xb2 (10...Af8 11 ±e3
^.e7 12 ^.d3 gives White a slight advantage,
Movsesian-Rogers, Prague 1999) 11 ^.xg7 (11
Sbl gxh6 12 #d2 £>xd4 13 Sxb2 £>xf3+ 14
Axf3 #c7 looks OK for Black) ll...Sg8 12
±f6 ±xal 13 #xal £>e7 14 ffcl £>f5 15 ±d3
a5 16 Axf5 exf5 17 e6! i.xe6 18 #h6 Sc8 19
#xh7 <4>d7 20 h4 oo Movsesian-K.Muller,
Austrian Cht 2000.
8 cxd4 £>ge7 9 £>c3 £>f5 (D)
W
AB BAHAO
Iff «HttrMl"[|g|
if
m.
10£ia4

5..±d7
65
After 10 Ae3 «xb2 11 ^b5 Sc8 12 Sbl
«xa2 13 Sal #b2 14 Sbl White can repeat the
position, and I don't think he can successfully
deviate.
10...Wa5 (D)
White's space advantage confers the better
chances upon him if Black plays passively with
10...Wc7 or 10...Wd8.
11 Ad2 Ab4 12 ±xb4
There is a similar line without 0-0 and ..JLd7
when White plays Ac3, but here 12 Ac3 can
simply be met by 12...^cxd4oreven 12...^xe5.
12...«xb4 13 a3 «e7 14 Scl
14 b4 could appear as an oversight, but in
fact White has very decent compensation in the
event of 14...£>cxd4 15 £>xd4 £>xd4 16 fxd4
Axa4 17 Sfcl, since White takes control of the
c-file.
14...b6!?
14...0-0 15 £>c5 ±.
15 £>c3 0-0 16 Ad3 £>h4
The game is approximately equal, Pyhala-
Tuominen, Helsinki 1992.
C2)
6...f6!? (D)
Another Watson-inspired idea.
7 0-0
This allows Black to fix a white pawn on e5,
which can later be attacked and should
consequently worry White since he might have to
spend time defending it. On the other hand,
maintaining the pawn on e5 keeps a valuable
m mwmmm
advantage in space, and makes Black's
development a little more problematic. Other moves
are worse:
a) 7 c4?! (with his king still in the centre,
White isn't ready for this) 7...cxd4! 8 cxd5
exd5 (8...Wa5+ 9 £>bd2 «xd5 10 ±c4 «c5 11
exf6 £>xf6 12 «e2 *) 9 exf6 £>xf6 10 £>xd4
Ac5 + Kupreichik-Dolmatov, USSR Ch
(Vilnius) 1980/1.
b) 7 exf6 £>xf6 8 0-0 is considered
relatively harmless. It is interesting to note that
after, for example, 8...cxd4 9 cxd4 Ad6 10 £ic3
0-0 (maybe 10...a6 is even more precise) we
have arrived at a type of position that occurs
much more often after 3 *hd2 *hf6 4 e5 Z£sfd7 5
Ad3 c5 6 c3 *hc6 7 $Se2 cxd4 8 cxd4f6 9 exf6
£hxf610 Qsf3 Ad611 0-0 0-012 *hc3, and now
Black usually goes 72... a6 (see Line A of
Chapter 7 in The French: Tarrasch Variation). Only,
in the line we are examining White's bishop is
more passively placed on e2 (rather than d3),
and Black has avoided White exchanging the
dark-squared bishops, which often gives more
cause for concern in the 3...^f6 Tarrasch
Variation.
We return to 7 0-0 (D).
Now:
C21: 7..J?b6 66
C22: 7...fxe5 68
7..."#c7 is feasible but not as forcing as
7...fxe5, and allows White an extra option in 8
Af4.
John Watson draws attention to a fourth idea,
one of his own, namely 7...cxd4 8 cxd4 fxe5 9

66
French: Advance and Other Lines
&xe5 (9 dxe5!?) 9...£>xe5 10 dxe5 Ac5. This
hasn't attracted much attention though, and thus
I shall give just a single line, which I believe is
safe and good for White: 11 ±d3 #h4!? (the
point of Watson's idea) 12 £>d2 &e7 (12...£>h6
13 £>f3 #h5 14 #e2 Ic8 15 ±e3 ±xe3 16
#xe3 0-0 17 h3 b6 18 £>d4 ± Ansell-S.Wil-
liams, Witley 1996) 13 £>f3 #h5 14 #e2 (14
Af4 0-0 15 ±g3 £>f5 16 Scl £>xg3 17 hxg3
JLb6, planning ...If7, ...Iaf8 and ...Wg4 with
an attack, is given by Watson but White's play
looks a little too cooperative) 14...0-0 15 JLe3
±xe3 16#xe3£>g6 17±xg6#xg6 18lacl ±.
C21)
7...#b6 (D)
W
This direct approach is linked with an
advance of the b-pawn, and is similar to 5...Jkd7 6
dxc5. White is aiming for better central control
by attacking Black's bishop and knight. Some
other moves may be worth considering too:
a) 8 exf6 £ixf6 9 b3 JLd6 was already very
comfortable for Black in Heyer-Read, Chicago
1991.
b) 8 Jtd3 (White loses a tempo with his
bishop but this isn't serious: its placement is
clearly improved and Black must watch out for
tactical problems on the kingside) 8...0-0-0!?
(if 8...cxd4 9 cxd4 £lb4 White can consider
both 10 JLe2 and 10 Sel!?, aiming for rapid
development; 8...fxe5 9 £lxe5 £lxe5 10 dxe5
0-0-0 11 c4!? might be better for White, since
Black lacks counterplay against e5 compared
with Line C12) 9 dxc5 ±xc5 10 b4 ±e7 11
±e3 (11 Af4!?) ll...Wc7 12 exf6 £>xf6 13 b5
£la5 14 JLd4 JLc5 °° Mukhametov-Beliavsky,
Budva 1996.
c) 8a3!?(Djandnow:
jab a a
This could also arise from other move-orders,
such as 5..Mb6 6 ke2 k.d7 7 0-0/6.
8 dxc5!?
cl) 8...c4 9 £>bd2 £>a5 - 5..Mb6 6a3c4 7
*hbd2 £ha5 8 k.e2 kd7 9 0-0/6.
c2) 8...fxe5 9 dxe5 (9 £>xe5 £>xe5 10 dxe5
and now 10...0-0-0 11 c4 is again a type of
position that occurs in Line C22 though Black's
queen appears misplaced on b6; Psakhis gives
10...£>e7! =) 9...c4 10 £>bd2 £>a5 11 £>d4
JLc5, Jonkman-Stellwagen, Wijk aan Zee 2003,
and now Psakhis gives 12 Ebl!?, preparing
b4; e.g., 12...£>e7 (12...±xd4 13 cxd4 #xd4
14 £>xc4! #xdl 15 £>d6+ ±) 13 b4 cxb3 14
£>2xb3 ±.

5...±d7
67
d) 8 £>a3! ? (this is less investigated than the
main line but might be White's best) 8...fxe5 9
dxe5 £>h6 (9...£>ge7 10lei £>g6 11 Ad3 0-0-0
12 c4! d4 13 £>g5 £>gxe5 14 Axhl Ae7 15 Ae4
Axg5 16 ±xg5 Ide8 17 b3 #c7 18 Af4 ±
Weijers-S.Weide, Hengelo 2000; 9...c4 10 b3
±) 10 c4! d4 11 Ad3 £>f7 12 lei Ae7 13 h4! (a
very strong move, preventing ...g5 for a while,
and thus giving White time to prepare an attack
on the queenside) 13...0-0-0 (13...&M!?) 14
£>c2 Sdg8 15 a3 g5 16 hxg5 £>xg5 17 Axg5
Jtxg5 18 b4, Grishchuk-Kolev, Istanbul OL
2000, and now Psakhis gives 18...Sg7 19 Ae4
lhg8 20 g3 Af4! 21 4>fl "with a complicated
position".
8...Axc5 9 b4 Ae7 (D)
10Af4
White probably does best to try to maintain
his control over e5. Other moves:
a) 10exf6Axf611b5 4ia5 12Ae3Wd813
Ad4 £Mi6! was fine for Black in Liepold-
Muehl, 2nd Bundesliga 1988/9.
b) 10 c4!? is interesting, and certainly
deserves attention: 10...dxc4(10...£ixe5!? 11 cxd5
#xb4 * Psakhis) 11 £>a3 £>xe5 12 £>xe5 fxe5
13 £>xc4 #xb4 14 a4 £>f6 15 £>xe5 #d6 16
£>xd7 #xdl 17 Ixdl £>xd7 18 Ab2 Af6 and
Black maintained equality in Degraeve-Bricard,
French Ch (Val d'Isere) 2002.
10...fxe5
10...£tfi6!? is a perfectly playable
alternative, aiming to step up the attack against e5 with
...£M7. Black seems to be doing well; e.g., 11
c4!? £>f7 12 cxd5 £>cxe5 13 a3 £>xf3+ 14
Axf3 e5 15 Ae3 #d8 16 £>c3 ^d6 = Mariotti-
Korchnoi, Thessaloniki OL 1988.
11 £>xe5 £>xe5 12 ±xe5 £>f6 (D)
It is time to complete development. Worse is
12...£f6?! 13 Ad3! ±xe5 14 #h5+ *e7 15
#xe5 £>f6 16 ^d2 Ihf8 17 lael *f7 18 Ze3
*g8 19 h3 with the better game for White,
Reefat-Leer Salvesen, Gausdal 2001.
13 a4
With this White decides to grab space on the
queenside before Black begins undermining
with ...a5. Another interesting idea is to force a
weakening of Black's kingside with 13 Ah5+
g6 14 JLf3, which worked well for White in
Grau-Leconte, corr. 1994: 14...0-0 15 lei Ad6
16 Axd6 #xd6 17 #d4 a5 18 ^d2 axb4 19
cxb4 with a slightly better position for White.
13...a5 14 b5
"There is nothing obviously wrong with this,"
remarks Kosten. Yet in J.Polgar-Topalov, Dos
Hermanas 1994, White chose 14 bxa5 2xa5 15
^d2 but was worse after 15...0-0 16 Ibl #a7
17 Ab5 Axb5 18 axb5 Ia2 19 Ad4 #b8! 20
Sel (Topalov's suggestion, 20 #e2!?, is maybe
better) 20...Wf4 21 Ie2 Ad6 22 g3 Wf5. Polgar
was probably concerned about her backward
c-pawn, and did not fancy closing the
queenside, and thus having no real counterplay on
that side if a c4 push fails.
14...0-015£>d2lac8(T>)
16 c4
Or:
a) 16 *hl and now 16...&d6? 17 £>c4! ±
is a nice little tactic, which caught Black in

68
FRENCH: ADVANCE AND OTHER LINES
W
V/,
I
s^^
AAA.
m
X
|B? ^1
, m mm* mm
I&iAi 1
A
%z
is a*ff
Sapab Pi
• HP A HP HP
iH JH *iH ill
IIP ?XP A ii_P IIP
W////^ / W////^ W////,y W///A
K.Lie-Leer Salvesen, Gausdal 2001. Nothing
seems wrong with 16..JLe8, after which 17
Ad3 can be met by 17.. JLd6.
b) 16 £>b3 Ad6 17 ttd4 «c7 18 ±xd6
«xd6 19 £>xa5 e5 20 «e3 b6 21 ^b3 £>e4 with
counterplay.
16...±e8 17 cxd5 £>xd5 18 £>c4 «d8 19
ttd2 4.b4 20ttb2«e7
Black is very active, and this compensates
for his inferior pawn-structure, Lacasa-Topalov,
St Cugat 1992.
C22)
7...fxe5
This straightforward, ambitious, but also
risky line has received a fair amount of
attention lately.
8£>xe5
Or 8 dxe5 «c7 9 Sel 0-0-0 10 Ad3 £>h6 11
£>a3 a6 12 Ag5 Se8 13 Af4 Ae7 14 Wd2 Sef8
15 £>c2 c4 16 Ml Shg8 with counterplay,
Movsesian-Shirov, Bundesliga 1999/00.
Generally, in these types of positions White tends to
prefer to exchange the knights on e5 too.
8...£>xe5 9 dxe5 (D)
9...«c7
This leads to very sharp play. Black is
hoping to gain time by attacking the e5-pawn and
prepares to castle queenside, but his centre is
very fragile and White inevitably obtains a
strong initiative.
9..JLc6 is an attempt to play the position in a
quieter manner. Black plans ...Wdl and ...£te7
with a more solid position but White has good
chances of an advantage due to his lead in
development: 10 c4 £te7 (this could also arise from
9...£he710 c4 Ac6) 11 Ag5 Wd7 (1 l...dxc4 12
£>d2! «d5 13 Ah5+ £>g6 14 «g4 was very
good for White in Benjamin-Gulko, Groningen
PCA 1993) 12 Ari5+, and now:
a) 12...g6 13 Af6 Sg8 14 Af3 dxc4 15
±xc6 £>xc6 16 £>a3 «xdl 17 Sfxdl £>d4 18
£lxc4 gave White an advantage in Keitling-
haus-Reefschlager, Bad Worishofen 1997.
b) 12...£>g6 13 1^4 (this looks stronger
than Benjamin's suggestion 13 ®c2, when he
gave 13...1T7 14 cxd5 ±, but Black can
improve with 13...d4, when he chances of equality
after both 14 ±xg6+ hxg6 15 «xg6+ Wf7 16
«xf7+ 4xf7 and 14 !ti3 Wf7 15 £>d2 Ae7)
13...dxc4(13...d4 14^d2±) 14<&c3 Wf7 15 f4
± S.B.Hansen-Peric, Cordoba 1994.
10 c4 (D)
\4
m.
mk
WM W
■^™%^;-"-^» war/,
m*m
« mm
y wm, mm
§ s
iAH
^^isy
10...0-0-0

5..Ad7
69
The immediate capture on e5 is regarded as
riskier due to the following fine effort by Rom-
anishin: 10...1^5 11 Ah5+! (White has no di- W
rect way of exploiting the following weakening
of Black's kingside but it is at least very
annoying for Black) ll...g6 12 Af3 0-0-0 13 Sel
#d6 (13...#f5 is possibly a slight
improvement but the same plan as in the game, viz. 14
£>c3 dxc4 15 #e2, promises White the better
chances) 14 £>c3 (White could win the pawn
back on d5 but this would only help Black
develop, so he instead strengthens the pressure
and hopes for 14...±c6? 15 Sxe6! #xe6 16
Ag4 +-) 14...dxc4 (14...d4?! 15 £>e4 «b6 16
b4! opens the queenside decisively) 15 #e2 £>f6
16 #xc4 Ae7 17 &f4 #d4? (this loses at once;
17..."#b6 18 b4! is given by Romanishin with
the pretty idea lS.-.^xr^? 19 Sabl! +-, while
17...»a6 18 #xa6 bxa6 19 Sadl is positionally
horrible but might be Black's best chance) 18
£>b5! AxbS 19 &xb7+ 1-0 Romanishin-Ivan-
chuk, Irkutsk 1986.
It is also possible to close the centre with
10...d4, though I find this slightly suspect
positionally since White easily blockades the
d-pawn, gets access to the e4-square and can
attack Black's queenside with a later b4. After 11
Af4 (11 Sel!?) 11...0-0-0 12 £>d2 £c6 Black
obtained a strong attack following 13 Ad3 g5! ?
14 Ag3 (it isn't clear what happens on 14
Jtxg5!?, which is a possible improvement)
14...h5 15 h3 g4 16 hxg4 h4! 17 £h2 h3 in
Jonkman-Smeets, Groningen 2002. White can
improve on this, of course, and the simplest is
probably Psakhis's suggestion 13 Af3! £>e7 14
Ae4^g6 15Ag3±.
We now return to 10...0-0-0 (D): W
11 Sel
This may be a slight refinement in the move-
order, as it avoids the older 11 cxd5 #xe5 12
JLf3, when Black, apart from transposing to the
main line with 12...exd5 13 Bel #d6, can try
12... Ad6. This can lead to exciting play after 13
g3:
& a) 13...£>f6 14 &f4! (14 £>c3 ff5 is equal)
14...1rxb2 15 Axd6 ©xal is a position
examined by Chytilek that gives White excellent
compensation. White's best continuation is 16
#d2! Ab5 (16...£>e8 17 Ae7 is also much
IM
AHif
■ A
wh «s*..« wm
m BAB
w, Hi
AH
*>m
1AH
""§nw
M««i«*
better for White) 17 Be 1! Bxd6 18 £>c3 #xc 1+
19»xcl±.
b) 13...#f5!? 14 £>a3 exd5 15 ±xd5 Ae6
16 ±xe6+ #xe6 17 #c2 is slightly better for
White.
c) 13...exd5 14 Af4 #f6 15 £>c3 ±c6 (or
15...£>e7 16 Axd6 #xd6 17 £>e4 followed by
£>xc5 ±) 16 £>xd5 #d4, Sveshnikov-Psakhis,
Sochi 1984. Now Sveshnikov gives 17 #c2! as
good for White but even better might be 17
Axd6 Bxd6 and only then 18 #c2.
Il...ttze5 (D)
Black can also refrain from taking the e5-
pawn by closing the centre with 11 ...d4, which
is similar to 10...d4. Lyell-N.Pert, Hastings
1995/6 continued 12 &f4 £>e7 13 Ad3 h6 14
£>d2 g5 15 Ag3 Ag7 16 a3 Bdf8 17 b4 h5 18
h3, intending £te4 with a clear advantage to
White.
ii§ W^XW^''''' IP
f|f 4
A
v,
'M
W/,p> Wflk, WB,, WOtik
vtT/, W//,
%nm
12 i.f3 0d6 13 cxd5 exd5 14 b4

70
French: Advance and Other Lines
This is a very logical idea; White is prepared
to sacrifice another pawn in order to open the
queenside. Other moves have promised little:
a) 14 g3! ? (a typical move in this line; White
prepares Jtf4 but also weakens his kingside)
14...£rf6 15 ±f4 Wa6 16 £>c3 d4 17 £>e4 £>xe4
18 Sxe4 Jtd6 19 Scl d3! with counterplay,
Sveshnikov-Vaiser, Sochi 1984.
b) 14 £>c3 £tf6 and then:
bl) 15 £>xd5 £>xd5 16 Wxd5 Wxd5 17 ±xd5
±f5 18 ±f3 ±d6 19 ±e3 V2-V2 Movsesian-
Shirov, Istanbul OL 2000.
b2) 15 b4 ±c6 16 bxc5 Wxc5 17 ±d2 <4>b8
18 Icl Wa3 19 Ibl ±d6 20 £>b5 ±xb5 21
lxb5 lfxa2 22 ±g5 and White might have some
compensation, but not more than that, Timman-
P.Nikolic, Dutch Ch (Rotterdam) 1997.
We now return to 14 b4 (D):
« ill
Wi
m m
iili 111
A ^.5.^ A A
1
■ m
/'"" 'wm>. ''Mat. j>, ''M
A" *
w,
W% ft M
14...£tf6
Most players would probably refrain
intuitively from taking on b4, but 14...Wf6 needs
calculation; e.g., 15&a3!?#xal 16^X05^6,
and now White has a draw with 17 We6+ $Ld7
18 Wd5 but he can also play for more by 17
Ag4+ Id7 (17... Adl 18 Axd7+ Hxd7 19 Ie8+
<4>c7 20 £sb5+ <4>b6 21 He6+ <4>xb5 221^7+
+-) 18 Wc4! (18 ±xd7+? ±xd7 19 £ib5 £>f6
20 £>xa7+ <A>b8 -+) 18...£>f6 (18...£se7!?) 19
£>c2 (19 b5? Ad5! -+) ^...Wbl 20 ±xd7+
±xd7 21 i.f411)2 22 ±g3! <A>d8 23 1T4 £>d5
24 1^8+ ±c8 25 £le3! with a winning attack
for White.
15 bxc5 WxcS 16 Ae3 Wa5 17 &d2 ±c618
a4!? ±d6 19 £sb3 fc7 20 g3 fihe8 21 flcl
White has promising attacking chances,
Timman-Brynell, Bundesliga 2000/1.
C3)
6...Hc8 (D)
w
ftW§ iff & w§ ft Wm
This has become fairly popular recently.
Black develops his queenside before deciding
what to do with the kingside.
7 0-0 a6 (D)
It is still not clear how Black will continue.
Depending on White's set-up he is probably
going to play ...£}ge7-f5, ...£\h6-f5 or ...^ge7-g6.
W
8<4>hl
White responds with a waiting move himself.
The position is still rather unexplored and this
may turn out not to be the most critical move but
it is very practical. So far, White has not been
able to prove an advantage with other moves:

5...±d7 71
a) 8 lei £>h6 9 dxc5 (9 ±xh6 gxh6 10 Wd2
deserves attention) 9... Jtxc5 10 Jtxh6 gxh6 11
#d2 #b6 12 ±d3 ±f8 13 #e2 ±g7 14 £>bd2
0-0 15 £>b3 = Grosar-Ekstrom, Pula Z 2000.
b) 8 b3 £>ge7 9 ±b2 £tf5 10 ±d3 cxd4 11
A.xf5 exf5 12 £>xd4 ±e7 13 lei 0-0 14 #d3
£>xd4 15 cxd4 #a5 = Sveshnikov-Epishin,
Moscow 1992.
c) 8 a3 cxd4 9 cxd4 £>ge7 10 £>c3 £tf5 11
A.e3 (11 g4 £>h4 12 £>xh4 #xh4 13 ±e3 g5!
followed by ...h5 gave Black a strong attack in
Xie Jun-Karpov, Guanzhou rpd 2000; then
White should probably try 14 f4!? with a
complex position) ll...±e7 12 ±d3 ^ixe3 13 fxe3
0-0 14 If2 f6 15 Wc2 f5 16 *hl We8 = Ivano-
vic-Bareev, Moscow 1993.
d) 8 £>a3 £>ge7 (8...cxd4 9 cxd4 £>ge7
would be another normal reaction) 9 dxc5 £\g6
10 b4 ±e7 11 ±e3 0-0 12 Wd2 £>gxe5 13 £>xe5
£>xe5 14 f4 £>g6 15 £>c2 ±f6 = V2-V2 Eren-
burg-Psakhis, Israel 2002.
e) 8 dxc5 ±xc5 9 ±f4 £>ge7 10 ±d3 f5!?
11 exf6 gxf6 12 £>h4 0-0 13 #g4+ <4h8 14
#h5 f5 15 b4 ±b6 16 £>d2 ±c7 (not 16...e5?
17 ^df3!, when White wins after both 17...exf4
18 £>g5 <4>g7 19 Wxh7+ <4>f6 20 Wh6+ and
17...#e8 18 Wxe8 Icxe8 19 £>xe5) 17 ±e3 e5
with counterplay, Oll-Ivanchuk, Polanica Zdroj
1998.
f) 8 g3 h6 9 h4 Wc7 10 h5 £>ge7 11 lei
cxd4 12 cxd4 Wb6 13 g4 g6!? 14 hxg6 £>xg6
15 £>c3 Ig8 16 <4>hl ±e7 17 Igl ±g5 18 Ibl
00 Grishchuk-Bareev, Kallithea ECC 2002.
g) 8 £>bd2 cxd4 9 cxd4 £>ge7 10 £>b3 £>g6
11 ±d2 ±e7 12 Icl 0-0 = Jonkman-Zhukova,
Wijk aan Zee 2005.
We now return to 8 *hl (D):
8...cxd4
8...^ge7 9 dxc5 £>g6 10 ±e3 £>cxe5 11
£>xe5 £>xe5 12 b4 ±e7 13 £>d2 £>c6 14 f4 ±f6
15 Icl 0-0 16 £tf3 Ie8 17 a4 gave White some
initiative in Grishchuk-Radjabov, Dubai FIDE
GP 2002.
9 cxd4 £>ge7 10 £k3 £>g6
10...£tf5 11 g4 £>h4 12 ±g5 ±e7 13 ±xe7
#xe7 14 £>xh4 #xh4 15 f4 0-0 leads to an
unclear position.
11 ±d3 ±e7 12 ±e3 0-0 13 lei £>a5 14
Icl £k4 15 Ic2 £>xe3 16 fxe3 b5
The chances are roughly equal, Grishchuk-
Radjabov, Dubai FIDE GP 2002.
C4)
6...£>ge7 (D)
Now:
C41: 7 0-0 72
C42: 7^a3 74
A few minor options include:
a) 7 dxc5 £>g6 8 ±e3 £>cxe5 9 £>xe5 £>xe5
10 f4 £>c6 11 0-0 g6 12 £>d2 ±g7 13 £rf3 0-0
14 h4 b6 15 cxb6 axb6 16 h5 £>e7 17 hxg6 hxg6
= Kharlov-Kramnik, Sao Paulo 1991.
b) 7 h4 cxd4 8 cxd4 Wb6 (with the idea of
...£}f5; the immediate 8...£tf5?! is worse: 9 g4!
£>h6 10 ±xh6 gxh6 11 g5 #b6 12 #d2 hxg5
13 hxg5 £\a5 14 £\c3 ± Prie-Kindermann, Uzes
1990) 9 £>a3 £tf5 10 £>c2 £>b4 11 *f 1 h5 12

72
FRENCH: ADVANCE AND OTHER LINES
g3 Jtb5 = Movsesian-P.Nikolic, Bundesliga
2002/3.
C41)
7 0-0 (D)
7...£\g6
The most natural square for the knight is on
f5 but the general problem with this square is
that it isn't particularly safe, and in many lines
White can drive the knight away with g4 and
aim for a kingside initiative. The text-move
intends a quick ...f6 break. Other moves:
a) 7...cxd4 8 cxd4 £lf5 (the preliminary
exchange keeps the centre closed but the
drawback is that White's knight is allowed out to c3)
9 £\c3 (D) and now:
al) 9...#b6!? - 6..Mb6 7 0-0 cxd4 8 cxd4
thge7 9thc3£hf5.
a2) 9...±e7 10 g4! £\h4 11 £\xh4 ±xh4 12
±e3 0-0 13 f4 f6 14 exf6 ±xf6 15 Scl g6 16
#d2 ±g7 17 ±f3 (17 £\a4 ± Nunn) 17...Wa5
18 ±g2 <4>h8 19 *hl Sac8 20 h3 #b4 21 Sfdl
± Nunn-Anand, Munich 1991.
a3) 9...h5 10 ±e3 (10 ±g5 ±e7 11 ±xe7
#xe7 12 #d2 g5 13 £\b5 g4 14 £>el *f8 15
£\c2 a6 16 £\c3 &g7 17 Wf4 f6 18 exf6+ Wxf6
19 Sadl Saf8 «> Romero-Korchnoi, Wijk aan
Zee 1992) 10...±e7 11 ±d3 (L.Webb-Hanley,
London 1999) ll...£\xe3 12 fxe3 ±.
a4) 9...a6 10 Jte3 (staying flexible with 10
a3!? might be better) 10...±e7 11 ±d3 £\xe3
12 fxe3 0-0 13 e4 (13 a3 f5 =) 13...£\b4 14 exd5
exd5 15 ±bl ±e6 16 £>e2 £\c6 17 #b3 #d7
18 £>f4 g6 19 #e3 Sae8 20 ±c2 ±d8 V2-V2
Kharlov-Dokhoian, USSR Ch (Moscow) 1991.
a5) 9...Sc8 10 a3 a6 11 ±e3 (11 *hl!?)
IL.AtJ 12 ±d3 £\xe3 13 fxe3 0-0 (D) and
a51) 14e4dxe4 15±xe4f5 16exf6±xf6 =
Knaak.
a52) 14 Scl f5 15 exf6 ±xf6 leads to a
fairly standard position with Black having no
problems at all, because White cannot seem to
get an attack going, and due to the pressure
against d4 it is not easy to achieve the e4
advance. Black might even have slightly better
prospects due to the bishop-pair; for example,
16 *hl (16 £>e2!?) 16...g6! 17 ±bl ±g7 18
£\a4 (18 e4? £\xd4!) 18...Sc7 19 £\c5 ±c8 20
#a4 e5 ? Nunn-Luther, Hastings 1994/5.
a53) 14&e2^a515£\f4g616lre2£\b317
Sadl #b6 18 h4 with some attacking chances

5...±d7
73
for White. Benjamin-Rey, San Francisco 1998
continued 18...Jtb5, on which Baburin
suggests 19 h5!? ±xd3 20 Wxd3 g5 21 £>h3 h6 22
£jh2, and if 22...Wb5 23 £>g4 *g7, then 24
Zxf7+! Sxf7 25 Wg6+ *f8 26 Wxh6+ <4>e8 27
5}xg5 with a strong attack.
b) 7...£tf5 8 ±d3!? cxd4 (on 8...£fo4, the
simplest seems 9 ^ixh4 lfxh4 10 ±e3 cxd4 11
cxd4 Wd8 12 ^c3 with the better game for
White, Blatny-Ruxton, Oakham 1990) 9 ±xf5
exf5 10 £>xd4 ±e7 (D) and now:
W
* A 6 6 A A A
«bl
Il4il§i
IABAI
API ^ flAfl
bl) 11 Sel ±e6 12 ±e3 0-0 13 £>d2 £>xd4
(13...£fce5 14 £>xe6 fxe6 15 i.d4 £>c6 16 Sxe6
£>xd4 17 cxd4 Wd7 18 Sel ±) 14 ±xd4 Wd7
15 f4 Sfc8 with approximately equal play,
Timman-M.Gurevich, Bundesliga 1999/00.
b2) Hlrb3Wc7!?(ll...^xd412cxd4Wb6
13 Wxb6 axb6 14 £>c3 ±e6 15 a3 <4>d7 =
Reefat-Barsov, Dhaka 2001) 12#xd5 Wxe5 13
Wxe5 £>xe5 14 b3 ±f6 15 £>a3 0-0-0 16 £>ac2
g6 17 a4 a6 = Hendriks-M.Gurevich,
Amsterdam 2001.
c) 7...Sc8(D) and then:
cl) 8 £>a3 cxd4 9 cxd4 and now 9...£tf5 10
£>c2 Wb6 transposes to 7 *ha3 cxd4 8 cxd4 $Sf5
9 foc2 Wb610 0-0 ScS but in Shirov-Yusupov,
Prague rpd 2002, Black put the knight on g6,
arguing that against this set-up White's knight
isn't particularly well placed on c2: 9...£}g6 10
£>c2 ±e7 11 ±d3 0-0 12 Sel f6 13 ±xg6 hxg6
14 Wd3 f5 15 Wd2 f4 16 b4 g5 17 h3 ±e8 18
Jta3 Jtg6, and Black was doing fine.
c2) 8 dxc5!? £>g6 9 b4 ±e7 10 ±e3 £>cxe5
11 £>xe5 £>xe5 12 £>d2 (12 f4 £>g6 13 f5 exf5
W
14 Wxd5 Wc7 oo) 12...0-0 13 f4 £>g6 14 £>b3
(14 £tf3 is well met by 14...b6) 14...We8! 15
±d4 ±d8 16 ±d3 ±c7 17 f5 (17 Wd2 ±b5 =)
17...exf5 18 ±xf5 ±xf5 19 Sxf5 We4 20 Sf2
£}f4 21 lfg4 £ie6 and White must play
accurately to keep an equal position, Peng Xiao-
min-Yusupov, Elista OL 1998.
8±e3
Or 8 g3, and then:
a) 8...±e7 9 h4 (D).
BaI
kmkmkmk
v M*Wtk
*m m
I
Black's knight is now about to be pushed
away, and there is now a choice between
releasing the central tension by exchanging on d4, or
maintaining it, which allows White to capture
on c5 at a suitable moment:
al) 9...cxd4 10 cxd4 0-0 11 £>c3 (11 h5
£>h8 12 h6 g6 13 £>bd2 f6 14 exf6 ±xf6 15
£ib3 §M1 16 £>h2 ^d6 with approximately
equal play, Hjartarson-Korchnoi, Amsterdam
1991) ll...f6 12 exf6 ±xf6 13 ±e3 £>ge7 14

74
FRENCH: ADVANCE AND OTHER LINES
±d3 ±e8 15 £>g5 ±xg5 16 ±xg5 h6 17 ±e3
(17 ±xe7 #xe7 18 ±b5, Torre-Motwani,
Manila OL 1992, and now Torre and Gloria give
18...Ic8 19 Wd2 e5 with equality) 17...±g6
(17...e5 18 ±c2 <*>) 18 ±xg6 £>xg6 19 #g4
#f6 20 ladl £ige7 = Malaniuk-Motwani, Forli
1991.
a2) 9...0-0 10 h5 £>h8 11 h6 (11 dxc5 ±xc5
12 c4 f6! 13 h6 g6 14 cxd5 £>xe5 was fine for
Black in Timman-Ljubojevic, Amsterdam 1999)
Il...g612dxc5f6!?13exf6±xc5 14±g5^f7
15 ±h4 #c7 16 #d2 £>ce5 with good
compensation for Black, Timman-P.Nikolic,
Amsterdam 1999.
b) 8...f6!? 9 ±d3 (9 h4? cxd4 10 cxd4 fxe5
11 h5 e4 12 hxg6 exf3 13 ±d3 #f6 + Swathi-
N.Pert, British Ch (Torquay) 2002) 9...cxd4 10
cxd4 £ib4!? (10...fxe5!? might also be tested)
11 exf6 gxf6 (ll...^xd3!? 12 fxg7 ±xg7 13
#xd3 0-0 with compensation is also an idea) 12
±e2 ±g7 13 £>c3 0-0 14 h4 (14 lei #e7 15
±fl Wfl 16 ±g2 Sad8 17 ±e3 ±c8 18 #d2
£\c6 = Jonkman-M.Gurevich, Hoogeveen 2001)
14...^e7 15h5^f5 16±f4±e8 17g4^d6 18
#b3 £>c6 19 h6 ±h8 20 ±d3 ± Jonkman-
Bukal, Valle d'Aosta 2002.
We now return to the position after 8 Jte3
8...cxd4
Equally important is 8...Jte7. It is the usual
dilemma of whether it is better to exchange on
d4 (which allows White's knight out on c3), or
to keep the central tension (which gives White
the option of capturing on c5). Then:
a) 9 £>el 0-0 10 f4 #b6 11 #d2 cxd4 12
cxd4 f6 13 ££3 fxe5 14 fxe5 Iac8 15 £>c3 £>a5
= Kupreichik-PNikolic, Ljubljana/Rogaska
Slatina 1989.
b) 9 dxc5 #c7!? (9...^gxe5 10£>xe5 £>xe5
11 f4 £>c6 12 £>d2 0-0 13 ±d3 g6 14 #f3 ±f6
15 #f2 ±g7 16 ££3 ± Kharlov-Sakaev, Sao
Paulo 1991) 10 £>a3 £>cxe5 11 £>xe5 £>xe5
(ll...Wxe5?! 12 ±d4 #g5 13 g3 ± Sakaev) 12
£>b5 ±xb5 13 ±xb5+ £>c6 14 c4 Id8 15 cxd5,
Kharlov-Sakaev, Russian Cht (Podolsk) 1992,
and now 15...Sxd5 is equal according to Sakaev.
9 cxd4 f6
9...±e7 10 £>c3 0-0 11 ±d3 f6!? (Il...±e8
is maybe safer but 12 2c 1 f6 13 exf6 Jtxf6 14
#d2 <4>h8 15 ±bl Ic8 16 Ifel gave White an
edge in Sax-Korchnoi, Wijk aan Zee 1991) 12
exf6 Jtxf6 13 Jtxg6 hxg6, Thipsay-Ravi, India
1993, and now Ravi gives 14 #d2! followed by
lael ±.
10 exf6 (D)
10...gxf6
10...#xf6!? 11 £>c3 ±d6 12 g3 0-0 was also
fine for Black in Veroci-O.Moor, Biel 1995.
Il^c3±d612lela6
with approximately equal play, Seul-Mov-
sesian, Bundesliga 2000/1.
C42)
7 £>a3 (D)
7...cxd4
This exchange makes perfect sense now that
White's knight has taken a different route and

5...±d7
75
w
isn't able to go to c3. It is also possible to stay
flexible with 7...£\g6, planning a later ...f6
break and keeping open the option of how to
deal with the central tension. Then:
a) 8 £>c2 ±e7 9 0-0 (9 ±d3 c4 10 ±e2 f6 11
exf6 !,xf6 12 b3 cxb3 13 axb3 0-0 14 0-0 h6 15
Jtd3 gave White an edge in Adams-P.Nikolic,
New York rpd 1994; it is maybe not right to
close the position with ...c4) 9...0-0 10 Sbl (10
lei cxd4 11 cxd4 Ic8 12 ±d3 a6 13 Ie3 f5
14 exf6 ±xf6 15 ±xg6 hxg6 16 Ic3 Wb6 =
Dvoirys-Tukmakov, Geneva 2001) 10...a5 11
A.e3 cxd4 12 cxd4 Wb6 13 a3 a4 14 !,d3 £>a5
15 We2 £>b3 16 h4 flfc8 17 h5 «tf8 18 £>h2,
Bangiev-PNikolic, Lippstadt 1999, and now
Black should maybe try to force through an
exchange of the light-squared bishops by 18...Sa5
followed by ... Jtb5, with approximately equal
chances.
b) 8 h4!? cxd4 9 cxd4 ±b4+ 10 *fl h6!?
(preventing White from advancing his pawn all
the way) 11 £\c2 ±e7 (D) has developed into a
critical line.
Now:
bl) 12 h5 £tf8 13 Ih3 £>h7 14 ±d3 0-0 15
Hg3 4>h8 16 ±f4 f5! (a key element of Black's
defence; White is also playing with some risk
since the h-pawn may become weak, and can
now be attacked with ...Jte8) 17 exf6 £\xf6 and
now:
Ml) 18^h4±e8!19^g6+±xg620lxg6
!,d6! 21 ±xh6!? (21 !,xd6 Wxd6, intending
...e5, is good for Black) 21...gxh6 22 Wd2 If7
23 2xh6+ £\h7, and Black successfully
defends according to Psakhis.
bl2) 18 £>e5 £>xe5 19 ±xe5 (19 dxe5 £>e4!?
is fine for Black, but not 19...^xh5? 20 Wxh5
Sxf4 21 Sxg7! with a smashing attack for
White) 19...±e8 20 Ih3 ±d6! and White's
initiative was successfully neutralized in Movses-
ian-PNikolic, Istanbul OL 2000.
b2) 12±d3(D)andnow:
b21) 12...£tf8 13 Ih3 (White keeps more
options open by leaving the pawn on h4) 13...Wb6
14lg3g6(14...g5!?)15 4>gllc816b3^b417
£>xb4 ±xb4 18 ±e3 ±b5 19 ±xb5+ Wxb5 20
Icl Ic6 21 h5!? ± Ehlvest-Minasian, Ohrid Ech
2001.
b22) 12...0-0 13 Ih3 f5!? 14 Hg3 4>h7 15
h5 £>h8 16 £>gl fan 17 £>h3 i.h4 18 Ie3 £>g5
19 4>gl lc8 20 £tf4 £ie4 with approximate
equality, Shirov-M.Gurevich, French Cht 2004.
8 cxd4 GXS 9 £k2 (D)
9 0-0 was seen in Shirov-Topalov, Dos Her-
manas 1997 but came from the move-order

76
French: Advance and Other Lines
6..&h6 7 0-0 Zhf5 8 foa3 cxd4 9 cxd4. The
game went 9...±xa3 10 bxa3 ^6 11 ±e3
£>xe3 12 fxe3 ^HdS\, which according to Shirov
is satisfactory for Black. Black may, of course,
also transpose elsewhere with 9...^6 10 £k:2
-9^c2^b610 0-0.
Now:
C421: 9...i.e7 76
C422: 9...£>b4 77
C423: 9...«b6 79
Or:
a) 9...Sc8 100-01Srb6-9...^6100-02c<3.
b) 9...Wa5+ 10 ±d2 Wb6 11 Ac3 Ae7 (Black
could also try 1 l...a5!? with the idea of 12 0-0
J.b4!?; perhaps White should prevent ...J.b4
with 12 a3) 12 0-0 a5 and now:
bl) 13 g4 £>h6 14 £tfel f6 (14...f5!?) 15
£>d3 £tf7 16 exf6 (16 Af37! fxe5 17 dxe5 0-0 is
slightly better for Black, Cifuentes-I.Sokolov,
Dutch Ch (Amsterdam) 1994) 16...gxf6 17
£>c5!? £>xd4 18 £>xd7 £>xe2+ 19 Wxe2 <A>xd7
20 Sadl followed by Sfel with some
compensation for White is given by I.Sokolov but I
don't quite trust this whole idea for White.
b2) 13 a3 0-0 14 ±d3 <A>h8 15 Sel £>a7 16
a4 £>c6 17 Ab5 and now 17...Ae8 18 Wd2 f6
19 exf6 Sxf6 20 £>e3 £>xe3 21 Sxe3 favoured
White in Timman-I.Sokolov, Dutch Ch playoff,
Amsterdam (3) 1996. Timman gives 17...f6!?
18 g4 (18 exf6 ±xf6) 18...£>h6 19 exf6 gxf6 oo
as a better idea for Black.
b3) 13 £\e3!? £\xe3 14 fxe3 0-0 15 ±d3
£>b4 16 ±bl a4 17 a3 £>c6 18 Wei £>a5 19
±b4 Wd8 20 ±d3 ± Galdunts-Bricard, Bisch-
willer 1999.
C421)
9...±e7 (D)
10 0-0
The sharper move 10 g4 is risky for White:
10...£\h4 11 £>xh4 ±xh4 12 0-0 f5!? 13 f4 (on
13 gxf5 Black had probably intended to
sacrifice a pawn with 13...0-0!?) 13...fxg4 14 ±xg4
0-0 15 b3, Sveshnikov-Karer, Nova Gorica
1999, and now I would prefer Black's position
after 15...Wb6.
After castling White is ready to play g4.
10...g5
This is Black's latest attempt, which has
mainly been experimented with by Zakhare-
vich. It leads to slightly unusual and complex
positions. Other moves:
a) 10...0-0 11 g4!? (11 ±d3 f6 12 Axf5 exf5
13 Sel ±) ll...£>h4 12^xh4±xh4 13 f4f6 14
exf6 £.xf6 15 ±e3 g6 16 Wd2 Wb6 17 b3 Sac8
18 Sf3 Sc7 19 Safl Sfc8 oo Sandor-Enders,
Bundesliga 1995/6.
b) 10...h5 11 Sbl!? (11 £\e3 is also good)
ll...Wb6 (ll...g5 12 b4 g4 13 «Mel Wb6 14
J.e3 favours White according to Grishchuk) 12
b4! a5 13 a3 axb4 14 axb4 ± Grishchuk-Kaida-
nov, New York 2000.
ng4(D)
11 b4!? a6 12 a4 f6 °° is given by Zakhare-
vich, although this looks risky in view of 13 b5
followed by exf6 and £te5.

5...±d7
77
m a.
™&m
AW> 6 IP AS A
ll...^h6!?
1 l...£\g7 12 b4 a6 13 Sbl (13 a4 looks more
logical) 13...f5 (13...b5!?Zakharevich) 14exf6
Jtxf6 15 a4 h6, Dvoirys-Zakharevich, Russian
Ch 2001, and now 16 b5 axb5 17 axb5 £\e7 18
£>e5 is better for White.
12 h3
12 h4 £\xg4 13 hxg5 Sg8 14 *hl h6 15 £>gl
£}gxe5 16 dxe5 hxg5 is unclear according to
Zakharevich, and indeed Black's attack
accelerates very quickly after moves like ...Wc7,
...0-0-0 and ...f6 or ...f5.
12...f6 13 exf6 ±xf6 14 b4 £\f 7 15 b5 £\e7
16 ±d3 Wc7
Black has very good counterplay in a rather
messy position, Goloshchapov-Zakharevich, St
Petersburg 2002.
This idea occurs frequently in this line of the
Advance Variation. Black hopes to ease his
position by exchanging a pair of knights.
10 £\e3!?
White would rather exchange the knight for
the one on f5. He will then kick the b4-knight
back again and can then follow up by
advancing the pawns on the queenside.
Alternatives:
a) 10 £\xb4 ±xb4+ 11 ±d2 aims for a
slight edge in an endgame, arguing that White
has a little more space and his bishop is a little
stronger than its counterpart. Though very safe
for White, his chances of an advantage seem
greater in the other lines:
al) 11...Wa5 is maybe a slight inaccuracy in
view of 12 a3! ±xd2+ 13 Wxd2 Wxd2+ 14
<4xd2 when, in comparison with line 4a22\
White has the extra move a3 for free. Anand-
M.Gurevich, Manila IZ 1990 then continued
14...f6 15 Sacl £\e7 16 b4 <4d8 (in fact this
position may deserve closer attention as a major
point of White's idea is supposed to be 16...£lc6
17 b5 £\a5 18 <4>c3 a6 19 a4, followed by <4>b4,
but 19...<A>e7 with the idea 20 <4>b4 £\c4! 21
±xc4 dxc4 22 Sxc4 axb5 23 axb5 Sa2 looks
like definite counterplay) 17 ±d3 Sc8 18
2xc8+ £>xc8 19 g4!? and White was better.
a2) ll...Wb6 12 ±xb4 (12 a3 ±xd2+ 13
Wxd2 ±b5 = is Black's point) 12...Wxb4+ 13
Wd2 (D).
C422)
9...£\b4 (D)
W
I
lip IIP ^P IP ^^
&» A 111 * Ji *
a.a m^MMp €8
Now:
a21) Dreev played 13...a5!? a couple of
times, and it doesn't look worse than the direct

78
FRENCH: ADVANCE AND OTHER LINES
transition into the endgame. 14 #xb4 (14 a3
1136! 15 0-0 0-0 16 Ifcl Ifc8 17 ±d3 £>e7
was fine for Black in Friedman-Dreev, New
York 1991) 14...axb4 15 4>d2 £>e7, Peng Xiao-
min-Dreev, Lucerne Wcht 1993, and now White
should try 16 a3!?.
a22) 13...#xd2+ 144>xd2^e7! 15lhcl f6
16 Ic5 (the rook is not doing much on the c-file
anyway, so even without the helpful move a3 as
in 'al' above, White should consider
advancing the b-pawn to b4 immediately) 16...4^8!
(planning ...±e8-g6/h5) 17 ±d3 Ic8 18 lacl
(18 Sxc8+ 4>xc8 19 Icl+ 4>d8 20 4>e3, with
an equal endgame, is more appropriate
according to M.Gurevich and Zlotnik) 18...Ixc5 19
Ixc5 ±e8 20 £>el £>c6 21 exf6 gxf6 22 4>e3
4>e7, intending ...&d6 and ...e5. Black already
has a comfortable position, Sieiro-M.Gurevich,
Havana 1986.
b) 10 0-0 £>xc2 11 #xc2 (D) and then:
bl) ll...h5!? 12 ±d2 ±e7 13 ±d3 #b6
(13...g5!? 14 ±xf5 exf5 15 #b3 ±) 14 ±xf5
Ic8! (a significant improvement over 14...exf5
15 ±g5! ±xg5 16 £>xg5 #xd4 17 Ifdl #h4
18 #d2! with excellent compensation for White,
Sveshnikov-Dreev, St Petersburg Z 1993) 15
#b3 exf5 16 #xd5 (16 #xb6 axb6 17 Ifcl =)
16...±e6 17 #a5 #xa5 18 ±xa5 b6 19 ±d2
Jtd5 and Black's bishops proved adequate
compensation for the pawn in Yagupov-Dreev,
Rostov 1993.
b2) 1 l...#b6 12 #d3 Ic8 (D) and now:
b21) 13 ±d2 ±b4! 14 ±f4?! (White should
prefer 14 ±xb4 #xb4 15 a3 #b6 16 Ifcl,
although Black has no problems in this
position) 14...a6! 15 a4 £>e7!? 16 h4 ±a5 17 Ifcl
Ixcl+ 18 Jtxcl h6 V2-V2 Sveshnikov-Dreev,
Rostov 1993.
b22) 13 g4!? £>e7 14 ±e3 h5 15 h3 hxg4 16
hxg4 a6 17 Ifcl (17 a4!?) 17...Ixcl+ 18 Ixcl
±b5 19 #d2 ±xe2 20 #xe2 £>c6 21 4>g2 ±e7
= Grosar-Drazic, Bratto 1997.
b23) 13 a4!? ±b4 14 h4!? h6!? (one of
White's points is 14...^e7 15 #e3! followed
by Jtd3 with a promising attacking position) 15
h5 £>e7 16 £>d2 (16 #e3 £>f5 17 Wf4 ±e7! =)
16...0-0 17 £>b3 f5 (Rublevsky gives 17...^c6?
18 Wg3 4>h8 19 ±xh6! gxh6 20 Wf4 £>e7 21
#xh6+ 4>g8 22 g4! followed by ±d3! +-) 18
a5 #d8 19 ±d2 ± Sveshnikov-Rublevsky,
Russian Ch (Elista) 1994.
b3) Il...lc8 12#d3a6 13a4±b4 14±g5
#b6 15 h4 h6 16 ±d2 #a5 17 ±f4 Ig8!?
(17...^e7!? is another plausible move; the
following is typically creative play from Moro-
zevich) 18 g3 4>d8! 19 #dl (19 #b3 4>c7 20
±d3 g5! 00) 19...4>c7 20 ±d3 4>b6 (note the
route Black's king takes - it cannot go via b8 in
view of the capture on f5 followed by e6+ but
via b6 is safe enough; 20...g5? 21 hxg5 hxg5 22
£>xg5 £>xd4 23 ±h7! +-) 21 £>el g5 22 hxg5
hxg5 23 ±cl 4>a7 24 ^c2 *a8, Movsesian-
Morozevich, Sarajevo 2000. This last move
with the king was maybe not entirely necessary
but it ends a remarkable journey in safety. The
position is approximately balanced.
We now return to the position after 10 £\e3!?
(D):
10...^xe3

There is no need to postpone this exchange.
Other moves:
a) 10...£>h4!? 11 £>xh4 Wxh4 12 0-0 £>c6
13 f4 Wd8 14 f5 Wb6 15 Sf4 ±e7 16 f6 gxf6 17
exf6 ±d6 18 Sh4 0-0-0 « Sax-Kindermann,
Aruba 1992.
b) 10...2c8 is seen occasionally but after 11
0-0 Black usually plays 1 l...^ixe3 12 fxe3 ±e7
anyway. The move ...2c8 is not a bad move to
get in but Black would probably rather have
castled in this position. 13 a3 £ic6 (13...£ia6 14
±d3 ±c6 15 We2 0-0 16 b4 £>c7 17 b5 ±e8 18
a4 and White was better in Dvoirys-Visser,
Hoogeveen 2001) 14 b4 a6 15 ±d2 0-0 16 ±d3
f5 17 exf6 ±xf6 18 Wbl g6 19 Wb3 £>e7 20 e4
t Balashov-Jurek, Schwabisch Gmiind 2001.
11 fxe3
This aggressive capture poses Black more
problems than the conventional 11 Jtxe3.
Il...£e7 12 0-0
12 a3 £>c6 13 b4 a6 14 Hbl £>a7 15 a4 £>c6
16 ±d2 a5 17 b5 ^b4 gave White nothing in
Spassky-Korchnoi, Belgrade Ct (18) 1977.
12...0-0 13 ±d2 (D)
13...a5
This was first suggested by Brenninkmeijer
but doesn't look sufficient for equality. Other
moves haven't done better:
a) 13...Wb6? 14 a3 £>c6 15 b4 f5 16 exf6
±xf6 17 ±d3 £>e7 18 Wbl h6 (18...g6 is a
better defence) 19 a4 ± Sax-Brenninkmeijer,
Wijk aan Zee 1992.
b) 13...£>c6 14 ±d3 f6 15 exf6 ±xf6 16
±c3 Wb6 17 Hf2 £>e7 18 We2 ±e8?! 19 £>e5!
± Dvoirys-Wempe, Dieren 2002.
14 a3 £k6 15 ±d3 Wb6 16 ±c3 f5 17 exf6
±xf6 18 Wc2 h6 19 Hf 2
White is better, Peng Xiaomin-Kacheishvili,
Istanbul OL 2000.
C423)
9...Wb6 (D)
The most common. Black attempts to
dissuade White from developing his dark-squared
bishop and prevents White from improving the
position of his light-squared bishop, which
would be much better placed on d3. Only later
will Black decide how to handle play on the
queenside. Usually ...£\b4 is played in
conjunction with ...Sc8 but in some lines Black may
also consider ...£\a5, intending ...Jtb5.
10 0-0
Lately Shirov and Movsesian have also
looked for sharper ideas:

80
FRENCH: ADVANCE AND OTHER LINES
a) lOg4£\fe7 11O-Oh5(ll...a5-700-0a5
11 g4 Zhfel is not bad either) 12 h3 hxg4 13
hxg4 f6!? (Van Wely isn't a man who shies
away from a tactical battle, and true to
character he decides to take the bull by the horns;
13...2c8 and 13...a5 are other options, while
13...£\g6 14 JLe3 JLe7 is given as equal by Van
Wely) 14 JLd3 0-0-0 (according to Van Wely
Black should prevent White's next move with
14...a5!) 15 b4! £\xd4? (Shirov recommends
instead 15.. JLe8; another idea is 15...f5!? 16 a4
fxg4 17 £\g5 and now 17...£tf5 or 17...£\xd4
18 JLe3 £tf3+ 19 £\xf3 Wc7 with a messy
position, as given by Psakhis) 16 £\cxd4 fxe5 17
We2! e4 18 Af4! £>g6 (18...e5 19 2acl+ £\c6
20 £\xe5 Wxd4 21 £.a6! Wb6 22 £>xc6 £.xc6
23 b5 +-) 19 2acl+ ix6 20 £.g3 Ad6, Shirov-
Van Wely, Wijk aan Zee 2001, and now White's
simplest is 21 JLxe4! dxe4 22 Wxe4 4~.
b) 10h4!?(T>Jandnow:
bl) 10...f6!? 11 g4£>fxd4!?(ll...£\fe7 leads
to a more passive position so Black goes for an
interesting piece sacrifice that will net two
pawns and active play for the knight) 12 £\cxd4
(12 £\fxd4! looks stronger; e.g., 12...£\xd4 13
Wxd4 £.c5 14 Wf4 fxe5 15 Wxe5 £.xf2+16 *dl
0-0 17 h5 with a powerful attack) 12...£\xe5!?
(12..JLM+!? 13 *f 1 £>xe5 may be even better)
13 g5 Ac5 14 0-0 £\xf3+ 15 fcxfi Wb4!,
planning ...Wg4+. Black has definite counterplay,
Movsesian-M.Gurevich, Sarajevo 2000.
b2) 10...&M 11 £>xb4 i.xb4+ (the
alternative ll...*xb4+!? 12 *fl £.b5 13 g4 £.xe2+
14 <&xe2 Wc4+ 15 Wd3 £>e7 16 h5 h6 17 £.d2
Sc8 18 a3 <&d7 19 Sacl Wxd3+ 20 <&xd3 Sxcl
21 Sxcl £\c6 leads to equality, Movsesian-
Ulybin, Croatian Cht (Pula) 1999) 12 *fl h5
(in Movsesian-Borovikov, Panormo Ech rpd
2002 and later in Movsesian-Berelovich, Bun-
desliga 2002/3 Black tried 12...£.b5 13 g4
Axe2+ 14 <&xe2 Wa6+ 15 Wd3 Wxd3+ 16
<&xd3 £\e7 17 h5 h6 but White was able to exert
some pressure in the endgame due to his space
advantage) 13 a4!? (an odd-looking move but
quite normal for this line since White wants to
preserve his light-squared bishop and argues
that Black can make little use of the b4-square)
13...Sc8 14 g3 i.e7 15 <&g2 Sc7 and Black is
doing fine, Shirov-M.Gurevich, Prague rpd
2002.
We now return to 10 0-0 (D):
10...a5
It is a game of shadow-boxing around here.
Black in particular has several ideas to choose
from, but right now he waits for White to
commit himself, but of course also makes sure that
White can't gain space on the queenside with
b4, and also in some cases prepares ...^b4 with
the intention of recapturing with the a-pawn.
The text-move also prevents White from playing
11 JLd3 because this can be met by 11 ...^cxd4
12 £\fxd4 £\xd4 13 £.e3 £.c5, when White
doesn't have 14 b4.
White's plan is clearer. In the majority of
cases he will try to kick the knight away from f5
with g4. This, however, will leave the kingside
open for a counterattack. White will continue by
moving the knight away from f3, so the f-pawn

5...±d7
81
can be advanced. This means either £\h4 or
£\fel must be played. The difference between
the two isn't great but there is some. Both
moves have the same intention of advancing
the f-pawn, but £\fel can often be met by ...h5,
sacrificing a pawn but regaining control of f5.
On £lh4 Black doesn't have the same reply
because the knight controls f5, and the situation
after a gxh5 is then much different. In that case
Black would often play something like ...£\g6
and ...f6 but this would require White to move
his knight again (£\g2), while this wouldn't be
necessary with the same plan against £\fel.
First we look at a number of alternatives for
Black:
a) 10...£>b4 11 £>e3 £>xe3 12 £.xe3 Sc8 13
£>el £.b5 14 a3 £.xe2 15 Wxe2 £>c6 16 Scl
£.e7 17 Sc3 0-0 18 Wc2 Wd8 19 £>f3 £>a5 20
£>d2 f5 21 exf6 £.xf6 22 Scl Sxc3 23 Wxc3
£\c6 24 £\f3 ± Vescovi-Dumont, Brazilian Ch
(Teresina) 2000.
b) 10...£>a5 11 g4 £\e7 12 £>fel £.b5 13
£>d3 h5 14 gxh5 £>f5 15 £.e3 £>c4!? 16 a4
(after 16 b3, 16...£\cxe3 17 fxe3 is better for
White according to Elwert but 16...£\a3 might
be better) 16...£>cxe3 17 fxe3 £.c4 18 £>f4 £.b3
19 £.b5+ <&d8 20 We2 £.xc2 21 a5! Wc7 22
Sfcl Sc8 23 a6b6 24 £.a4 Wc4 25 Sxc2 Wxe2,
Sveshnikov-Dolmatov, USSR Cht
(Naberezhnye Chelny) 1988, and now 26 Sxe2 JLe7 27
Sg2! Sc4 28 <&f2 (Sveshnikov) gives White a
clear advantage.
c) 10...£.e7 11 Sbl (11 g4 £>h4 12 £>xh4
JLxh4 13 g5 should be investigated, or just 13
£.e3) ll...a5 (Korchnoi gives ll...h5! 12 M a5
as the right defence for Black) 12 g4 £\h4 13
£>xh4 £.xh4 14 £.e3 (14 g5!?) 14...£.e7 15 f4
£>b4 16 a3 £>xc2 17 Wxc2 Sc8 (17...f6!?) 18
Wd3!? 0-0 19 *hl, Adams-Korchnoi, Garmisch
rpd 1994, and now Korchnoi gives 19...f6 20 f5
a4 as unclear.
d) 10...h5 11 £.d3!? (11 M!? might give
White some play for the pawn after 1 l...£\xb4
12 £>xb4 Wxb4 13 £.g5 but Black can also
decline the sacrifice, with 11...a5 for example,
also with complex play; in Shirov-Korchnoi,
Drammen 2004/5 Black was also doing very
well after 11 b3 a5 12 £.b2 £.e7 13 £>e3 £>xe3
14fxe3 a4 15 Sbl axb3 16 axb3 Sa2) 1 l...£>h6
(this retreat might not be entirely necessary,
and ll...a5 instead looks natural, even if 12
JLxf5 exf5 13 £>fel gives White a tiny edge) 12
M a5 13 bxa5 £\xa5 14 JLg5 with a substantial
advantage for White, McDonald-N.Pert,
Oakham 2001.
e) 10...Sc8 (D) and now:
el) 11 g4 £>fe7 12 £>h4 (12 £>fel h5 13
gxh5 £>f5 14 £.e3 £>b4 15 £>xb4 £.xb4 was
fine for Black in Ang.Perez-Jo.Medina, Cama-
juani 2000) 12...£>g6 13 £>g2 f6 14 exf6 gxf6
15 *hl JLd6 (Atalik gives 15..JLg7!, intending
16 f4 f5!) 16 f4 £>ce7 with dynamically
balanced chances, Sveshnikov-Atalik, Slovenian
Cht (Bled) 2000.
e2) 11 sfehl. This flexible move seems to
have established itself as a more dangerous
approach than the direct advance of the g-pawn.
Then:
e21) ll...£>b4 12 £>xb4 Wxb4 13 a4!
(preventing Black from exchanging bishops with
...£.b5) 13...a5 14 g4 £>e7 15 £.d2 Wb6 16 M!
axb4 17 a5 Wd8 18 £.xb4 h5 19 g5 gives White
the better prospects in view of his chances on the
queenside, Degraeve-Barsov, Montreal 2002.
e22) ll...£>a5 12 g4 £>e7 13 £>fel £.b5
(13...h5 14 gxh5 £>f5 should also be tested) 14
£>d3 h5!? 15 gxh5 £>f5 16 £.e3 g6?! (a strange
decision, even if it is difficult to foresee the
consequences; 16...£\c4 is stronger) 17 hxg6
fxg6 18 Sgl Wc7 19 £>del £.xe2 20 Wxe2
Wh7 21 £>f3 is much better for White, Grish-
chuk-Gulko, Esbjerg 2000.
We now return to 10...a5 (D):

82
French: Advance and Other Lines
11...H5
So White isn't getting a second chance to
play g4.
Il...£>b4 12£>e3! Sc8 (12...£>xe3!? 13 fxe3
JLe7, intending... Ab5, doesn't look bad; White
has the semi-open f-file but it isn't easy to
generate a real attack) 13 £>xf5 exf5 14 ±d2!? ±e7
(14...±b5!? M.Gurevich) 15 a3 £>c6 16 ±c3
(16 b4! is also very good, according to
M.Gurevich; e.g., 16...axb4 17 axb4 ±xb4? 18 Sbl
Wa5 19 Ixb4! £>xb4 20 Wb3 ±) 16...0-0 17
Wd2 ± Grishchuk-M.Gurevich, Esbjerg 2000.
12 b3
It is not essential to cover the b-pawn before
a) 12 JLe3 h5 13 h3 Hc8 14 Jtd3 hxg4 15 the dark-squared bishop is developed but it is
hxg4 £>a7 16 £>g5 (16 a4!?) 16...±b5 17 £>el always nice to know that it won't suddenly drop
£\ec6 ¥ Hendriks-Piket, Dutch Ch (Amster- off. There are also other options for White:
dam) 1996. a) 12 ±d3!? h4 13 ±xf5 exf5 14 h3 ±e7 15
b) 12 £>h4 f6!? (12...£>b4 13 ^a3 £>g6 14 ^e3? (15 b3 followed by ±g5 looks more to the
£>g2 f6 15 exf6 gxf6 16 f4 f5 17 h3 0-0-0 18 point) 15...±e6 = Ahmad-Barsov, Aden 2002.
±e3 4b8 19 £>bl ±b5 20 ±xb5 Wxb5 21 ^c3 b) 12 ±g5 ±e7 (12...Sc8 13 Wd2 ±e7 14
Wc4 gave Black counterplay in Jonkman-Blees, Sabl ±xg5 15 Wxg5 £>ce7 16 ±d3 ±b5 =

5..±d7
83
Peng Xiaomin-Zhang Zhong, Shenyang 1999)
13 ±xe7 £>cxe7 14 Wd3 Ic8 15 labl 0-0 16
£>e3 g6 17 h3 d?g7 with roughly equal play,
Dvoirys-Willemze, Dieren 2002.
12...±e7 13 ±d3 £>h4 (D)
Not 13...£>fxd4? 14 £>cxd4 £>xd4 15 ±e3
±c5 16 Hcl! +-, and 13...^b4 14 £>xb4 axb4
15 Axf5 exf5 16 ±g5 is also better for White.
14£>g5
14 £>xh4 ±xh4 15 ±e3 is a natural
alternative and White might also be slightly better
here.
14...a4
14...£>xd4? 15 £>xf7! £>xc2 (15...4xf7 16
£ixd4 ±) 16 £ixh8 £ixal 17 Wxh5+ 4d8 18
±g5 +-.
15 Ibl axb3 16 Ixb3!? Wd817 £ih3 Sxa2
18 Sxb7 £ia5 19 Sb2 Ixb2 20 ±xb2
w
Despite the concentration of forces on the
kingside and the rather symmetrical pawn-
structure, White has slightly the better of it,
Ehlvest-Barsov, Dhaka 2001.

5 5...^ge7 and 5...£Mi6
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 £>c6 5 £>f3 (DJ
Black's counterplay in the Advance
Variation is often centred around an attack against
d4. There are several ways to implement this,
with 5...Wb6 and the more flexible 5... JLd7
being the most common. These try to avoid a
white set-up with the bishop on d3, which in
general is the most dangerous type of set-up in
the Advance Variation.
A type of attack against d4, which we
discuss in this chapter and that often occurs also in
the two main lines, is the attempt to bring a
knight to f5. There are two routes: ...£\ge7-f5
or ...£Mi6-f5, both with specific drawbacks that
White can try to exploit.
First, if Black plays 5...£\ge7 White can
reply 6 £\a3 with the idea of playing £\c2 to
protect d4, and thus prepare a set-up with JLd3.
Additionally, the knight conveniently controls
the b4-square from c2. Hence, an increasingly
popular idea is to develop the knight via h6.
Then the £\a3-c2 idea is less appetizing with
Black being able to play ...JLxa3. The obvious
drawback to 5...£\h6 is that White may capture
the knight and double Black's pawns on the h-
file, although theory states that this is unlikely
to be very dangerous.
Finally, White might simply play JLd3 and
capture the knight when it comes to f5. This
damages Black's pawn-structure but the
position remains closed, and White is not in an
immediate position to exploit this.
We shall look at:
A: 5...£>ge7 84
B: 5...£>h6 87
A)
5...£>ge7 (D)
6£>a3
Like in other lines with an early ...£\ge7, this
is a good idea. White brings his knight to c2,
where it supports d4 and thus prepares a set-up
with JLd3.
Most other moves transpose elsewhere:
a) 6 ±d3cxd4 7 cxd4^f5-5...^h6 6 ±d3
cxd4 7cxd4%sf5.
b) 6±e2fof5-5...£hh6 6&e2thf5.
c) 6a3£hf5-5...£hh6 6a3£hf5.
d) 6 dxc5 £>g6 7 b4 £>cxe5 8 £>d4 £.e7 9 f4
£>c6 10 £.b5 £.d7 11 £.xc6 bxc6 12 0-0 0-0 13
±e3 £.f6 14 £>d2 Wcl 15 ®cl a5 16£>2f3 Sfe8
17 4hl Wbl 18 a3 Wa6 = Shilin-Nosenko,
Kharkov 2002.

5...£hge7 AND 5...Ehh6
85
6...cxd4
Often in the Advance Variation Black is
reluctant to make an early capture on d4, for it
might then be possible for White to develop his
knight to c3. Here White has already moved his
queen's knight and hence the capture becomes
more logical and indeed safer. It is now
advisable for Black not to delay this or White might
get additional possibilities later on. One
example is 6...£>f5 7 £>c2 Ml 8 £.d3 cxd4 9 £.xf5
exf5 10 £\fxd4!? with an edge for White, Svesh-
nikov-Dvoirys, Cheliabinsk 1989.
7cxd4£>f5 8£>c2(DJ
8...£.d7
There are a number of alternatives:
a) 8...£>h4 9 £>xh4 «xh4 10 £.e2 £.d7 11
0-0 f5 12 b4 !tt8 13 b5 £>a5 14 £>e3 favoured
White in Motylev-Potkin, Russian Cht (Togli-
atti) 2003.
b) 8...£.e7 9 £.d3 0-0?! (9...Wb6 - 8...Wb6
9 M3 M7) 10 g4! £>h4 11 £>xh4 £.xh4 12 g5
£.xg5 13 Wh5 h6 14 Igl £.xcl 15 Sxcl f5 16
Wxh6 Sf7 17 <A>e2 with an almost decisive
attack, Sveshnikov-Farago, Hastings 1984/5.
c) 8...»a5+9Ad2»b610Ac3Ae7(Black
has wasted a tempo with his queen but may
argue that White's bishop is poorly placed on c3;
on the other hand White now has b2 and d4 well
defended and can concentrate on building up an
attack on the kingside) 11 M3 a5 12 a3 Ml
13 0-0 0-0?! (Hodgson queries this and
suggests that Black should stop White's next with
13...h5) 14 g4 £>h4?! (14...£>h6 15 £>e3 f6!? is
more circumspect) 15 £\xh4 JLxh4 16 f4 with a
promising attacking position for White, Hodg-
son-Lalic, Port Erin 1995.
d) 8...£\b4 9 £\xb4 (recent games have seen
White opt for 9 Ml but it seems a rather
academic discussion whether it is more accurate
than the text-move; e.g., 9...£\xc2+ 10 Wxc2
Wa5+ 11 £.d2 £.b4 12 a3 Axd2+ 13 *xd2
1lxd2+14 4>xd2 leads to a very similar position
with a small endgame advantage for White,
Svidler-Luther, Bundesliga 2002/3) 9...JLxb4+
10 £.d2 «a5 11 £.d3 Axd2+ 12 «xd2 «xd2+
13 4xd2 £>e7 14 b4 £.d7 15 2nd ± Karpa-
chev-Lukov, Hyeres 2001.
e) 8...»b6 9 £.d3 (D) (9 £.e2 - 5..Mb6 6
M2 cxd4 7 cxd4 foge7 8 foa3 *hf5 9 *hc2) and
now:
A Hi A
el) 9...£.e7 10 g4!? (10 h4 £.d7 11 *fl -
8...M7 9 M3 ^b610 */7 M711 M\ 10 0-0
Ml =) 10...£>h4! 11 £>xh4 £.xh4 12 0-0 (12
g5? £>xe5!) 12...£.e7 13 £.e3 Ml 14 b4!? a5
15 bxa5 £\xa5 16 f4 Sc8 with an unclear
position, Plaskett-Mestel, Hastings 1986/7.
e2) 9...£.b4+ 10 *fl £.e7 11 h4 h5 (after
ll...JLd7?! White is a tempo up on the line
8...M7 9 M3 ^b6 10 */7 M7 11 h4 and
gained a clear advantage in Rohde-Spraggett,
New York 1986 after 12 g4 £>h6 13 Igl) 12 g3
a5 13 4g2 Ml 14 £.xf5 exf5 15 £.g5 ± Mal-
aniuk-Lputian, USSR Ch (Kiev) 1986.
e3) 9...£>b4 10 Axf5 exf5 11 0-0 ^c6?!
(Il...£>xc2 12 *xc2 Ml =) 12 Ibl h6 13
£>fel Ml 14 £.e3 a5 15 £>d3 ± Hendriks-
Dgebuadze, Dieren 2001.
9 £.d3 (D)

86
French: Advance and Other Lines
9 ±e2 - 5...kd7 6 ke2 foge7 7 tha3 cxd4 8
cxd4 &/5 9 *hc2.
I
9...Wb6
9...£>b4 10 £>xb4 £.xb4+ 11 £.d2 £.xd2+
(1 L..Wb6 12 £.xb4 Wxb4+ 13 Wd2 Wxd2+ 14
<4>xd2 is a little better for White, even though
Black has an extra move compared with note 'd'
to Black's 8th move) 12 Wxd2 «?b6 13 £.xf5
exf5 14 0-0 0-0 15 Hfcl is slightly more
comfortable for White, but probably not enough for a
serious advantage, Romero-Bukal, Rome 1986.
io*fi
An ambitious move that indirectly defends
d4, in view of 10...£>cxd4? 11 £>fxd4 £>xd4 12
JLe3 JLc5 13 b4 +-. With it, White hopes to be
able to avoid capturing on f5. Other moves:
a) 10 0-0 a5 (D) and now:
W
and, with a firm blockade of White's passed
centre pawn, no real worries about the isolated
d-pawn and dynamic possibilities on the king-
side, Black stands better, Jonkman-M.Gurevich,
Vlissingen 1997.
a2) 11 a3! ? a4 (Black's last move seemed to
prepare a capture of White's d-pawn, but Haba
points out that it is still premature to take the
pawn; e.g., 1 l...£>fxd4 12 £>fxd4 £>xd4 13 £.e3
£.c5 14 b4 axb4 15 Wg4! £>xc2 16 £.xc5 Wxc5
17 axb4 ®c8 18 Sxa8 ®xa8 19 Wxgl Sf8 20
£.xc2 ±) 12 Axf5 exf5 13 £>fel £>a5 14 £>d3
£>b3 15 Sbl £.e7 16 £>f4 ±e6 17 Ad2, Haba-
Kristjansson, Pardubice 2002, and now Haba
gives 17...£>xd2 18 Wxd2 0-0 19 £>e3 Wb5 as
equal.
b) 10 £.xf5 exf5 11 0-0 h6 12 a3 a5 13 lei
(this is certainly not imperative; the immediate
13 h4 looks more accurate, or M.Gurevich's
suggestion 13 Hbl!?, aiming for b4) 13...JLe6
14 h4 £.e7 15 g3 Wd8! 16 h5 g5 17 hxg6 fxg6
18 <A>g2 g5 19 Shi <A>d7! 20 Wd3 Wb6! and
again Black has managed a strong blockade on
e6 and achieved the ...g5 advance, with
promising play on the kingside, Blees-M.Gurevich,
Gent 1997.
10...£.e7 11 h4 (D)
11 g4?! £>h4 12 £>xh4 ±xh4 13 g5? is met
by 13...£\xe5. Hence White prepares to kick the
knight away from f5 without allowing ...£\h4.
W P
pi HH m|
al) 11 £xf5 exf5 12 Ae3 £>d8 13 2bl h6
14 h4 Ae7 15 h5 £>e6 16 g3 g5 17 hxg6 fxg6,
ll...f6!?
This leads to complicated play but it is very
logical to attempt to open the f-file against
White's king. Instead, from a slightly different

5..&ge7 AND 5...Zhh6
87
move-order, the game Movsesian-Lyrberg,
Neum ECC 2000 went 1 l...a5 12 Axf5 exf5 13
h5 f4 (13...h6 14 <A>gl £>b4 might be better,
when 15 £\e3 f4 gives Black counterplay) 14
Axf4Wxb2 15£>e3£>b4 16*gl fic8 17*h2!
(17 a3? is well answered by 17...&d3!) 17...h6
(17...#xf2!? isn't clear) 18 a3 Ag5 19 *g3!?
Jtxf4+ 20 <A)xf4 (a truly remarkable journey of
the white king - it will find safety on g3!)
20...0-0!? 21 <4>g3! (21 axb4? #xf2 22 £>xd5
g5+ 23 hxg6 fxg6+ 24 £>f6+ Ixf6+ 25 exf6
Ie8! and Black wins) 21...£>c6 22 #d3 with
the better position for White.
12 g4?!
Obviously, this is a critical move but 12
Jtxf5 exf5 13 h5 is more circumspect.
Alternatively, 12 exf6 ±xf6 13 g4 £>d6 14 g5 ±e7 also
looks fine for Black.
12...£ifxd4
12...fxe5!? is a serious alternative, which in
fact looks very good for Black in all lines: 13
£>xe5 £>xe5 14 dxe5 £>xh4; 13 dxe5 £>h6 14
Jtxh6 gxh6 15 4g2 0-0; or 13 gxf5 e4 14 fxe6
Axe6 15 £>g5 exd3 16 #xd3 Ad7.
13 £>cxd4
Or 13 £>fxd4 £>xd4 14 ±e3 fxe5 15 £>xd4
exd4 16 ±xd4 #d6 17 ±xg7 Sg8 18 ±c3 e5
19 g5 with an unclear position.
13...£>xe5!? 14 £>xe5 fxe5 15 £>c2 0-0
Black has good compensation for the piece,
Afek-Stellwagen, Wijk aan Zee 2000.
B)
5...£>h6 (D)
Now:
Bl: 6dxc5 87
B2: 6±d3 89
Minor alternatives:
a) 6 ±e2 £>f5 7 g4 £>h4 8 £>xh4 Wxh4 9
±e3 ±dl 10 £>d2 cxd4 11 £>f3!? #d8 12 cxd4
h5 13 gxh5 Sxh5 14 h4!? with roughly balanced
play, Kupreichik-Moskalenko, Lvov 1988.
b) 6 a3 and now 6...&f5 7 b4 cxd4 8 cxd4
leads elsewhere after 8.. Jtd7 - 5...±d7 6 a3
£hge7 7 b4 cxd4 8 cxd4 £j/5, or 8...#b6 -
5...Wb6 6 a3 *hh6 7 b4 cxd4 8 cxd4 Qtf5, but
Black also has other options, such as 6...a5 or
6...c4.
c) The immediate capture on h6 deserves
attention though: 6 Jtxh6 gxh6, and now 7 Jtd3
is an unexplored idea. In Afek-Vaganian, Dutch
Cht (Breda) 2001 Black reacted logically with
7...f6! 8 0-0 fxe5 9 £>xe5 £>xe5 10 dxe5 c4 11
±c2 #g5 12 f4 ±c5+ 13 *hl #g7 14 b4!?
cxb3 15 axb3 0-0 16 #e2 ±d7 17 b4 ±b6
(17...±e7!?) 18 c4 dxc4 19 Sa3 *h8 20 £>d2
If7 21 £>e4 Iaf8 22 £>f6 Ixf6 23 exf6 #xf6
24 #xc4 but the position remains unclear,
perhaps even in White's favour.
Bl)
6 dxc5(D)
It is very logical to try to exploit Black's last
move by simply capturing the knight. White's
move is very aggressive and indeed intends to
take on h6 but first he wants to lure Black's

88
FRENCH: ADVANCE AND OTHER LINES
bishop into play, so that h6 is undefended and
can be attacked later by Wd2. Furthermore,
White wants to seize better control of the
central squares by kicking Black's knight away
from c6 with b4-b5, which will now come with
gain of tempo.
6...£.xc5 (D)
6...£>g4? was once played by Bareev but is
more or less refuted by 7 Wa4!, which defends
the e5-pawn by pinning the c6-knight, and
attacks g4. White is then a pawn up for nothing
because Black's knight is going to return to h6
and White can then simply defend his pawn on
c5. Sveshnikov-Bareev, Poliot-T.Petrosian'
ECC 1991 went 7...h5 8 h3 £>h6 9 £.e3 £rf5 10
£.d4 £.d7 11 £.b5 ±.
7b4
It is worth noting the move-order. The text-
move is more accurate than 7 JLxh6 gxh6 8 b4
because Black can reply 8..JLf8! 9 b5 £>e7! 10
£.d3 £.g7 11 0-0 £>g6 as in Sveshnikov-Glek,
Moscow 1991. It is very useful for Black to
have his bishop on g7, where it defends h6 and
attacks e5.
7...£.b6
The idea from above, 7.. JLf8, would now be
less meaningful, since White then delays
capturing on h6, and simply continues 8 b5
(probably better than 8 £.d3 f6!) 8...£>a5 9 £.d3.
8 b5 £>e7 (D)
8...£>a5 is also feasible but White can then
go directly for the h6-pawn by 9 JLxh6 gxh6 10
Wd2 when, in Sveshnikov-Cherniaev, Moscow
1996, Black did not have enough compensation
after 10...£.d7 11 Wxh6 We7 12 £.d3 0-0-0 13
0-0 Sdf8 14 £>bd2.
KmjLmwm m
9£.d3
Sveshnikov turned to this move-order after
the game Sveshnikov-Nikolaev, Moscow 1992,
which went 9 JLxh6 gxh6 10 JLd3 (the naive
continuation 10 Wd2?! £>g6 11 Wxh6 Wc7 is
awkward for White) 10...£>g6 11 0-0 Well
(better than ll...f6? 12 £.xg6+ hxg6 13 Wd3
*H 14 exf6 Wxf6 15 £>bd2 Id8 16 lael,
which is much better for White, Sveshnikov-
Dukhov, Moscow 1992; 11...0-0 - 9 ±d3 ?hg6
10 0-0 0-0 11 kxh6 gxh6) 12 lei 0-0 13 a4
£.d7 14 Ia2 f6 (14...a6!? - Sveshnikov) 15
JLxg6 hxg6 16 Wd3 f5 with counterplay.
However, I haven't completely worked out if White
gains anything specifically from the change in
the move-order.
9...£>g4!?
An interesting attempt to punish White for
delaying the capture on h6. The logical
alternative is 9...£>g6 10 0-0 0-0 (on 10...1rc7, White
can try to avoid 11 JLxh6 gxh6 - 9 JLxh6 gxh6
10 ±d3 Zhg611 0-0 Wc7 but it is not clear that
there is anything particularly better; 10...^g4 -
9...Zhg4!? 10 0-0 Zhg6) 11 £.xh6 gxh6 12 a4 f6
13 £.xg6 hxg6 14 Wd3 <&g7 15 £>bd2 £.d7 16
c4 fxe5! 17 cxd5 exd5 18 Wxd5 Af5 19 Wxe5+
Wf6 with compensation, Sveshnikov-Moska-
lenko, Rostov 1993.
10 0-0 £>g6 11 £.xg6 fxg6 12 h3 £>h6 13
£.xh6 gxh6 14 £>bd2 0-0 15 c4 a6!
Black has counterplay, Grishchuk-Bareev,
Cannes rpd 2001.

5...*hge7 AND 5...Zhh6
89
B2)
6 &d3 (D)
6...cxd4
The exchange on d4 is good preparation for
...5M5, since White might be able to meet a
later exchange by favourably recapturing with
the knight; e.g., 6...£>f5 7 0-0 (7 £.xf5 exf5 8
dxc5 &xc5 9 0-0 &e6 10 £>bd2 d4 11 £>b3
&xb3 12 Wxb3 Wb6 seems OK for Black)
7...cxd4 8 Axf5 exf5 9 £ixd4 with an
advantage to White.
7cxd4
Here, 7 JLxh6 gxh6 8 cxd4 is noteworthy.
Black has been lured into an early capture on
d4, which may not be in his interests. On the
other hand, JLxh6 captures rarely cause Black
that much trouble. In Grishchuk-Bareev, Pan-
ormo ECC 2001, Black reacted too ambitiously
and was severely punished: 8....&d7 9 £ic3
Wb6 10 &b5 2g8?! (this was probably based
on an underestimation of White's chances in
the following complications; 10....&g7
intending ...0-0 and ...f6 looks right) 11 0-0 £ixe5?
(this is what Black prepared with his last move,
but it simply fails tactically) 12 £ixe5 JLxb5 13
Wh5 Sg7 14 Sfel (Bareev must of course have
seen this far, but perhaps did not investigate the
position carefully enough; White has an
overwhelming initiative, which causes Black's
position to collapse in only a few moves) 14...2d8?!
15 £>xb5 Wxb5 16 £>xf7! 2xf7 17 2xe6+ 1-0.
He loses one of his rooks after 17...^.e7 18
2xe7+! <&xe7 19 2el+ *f8 (or 19...<&d6 20
Wxf7) 20 Wxh6+ <&g8 21 Wg5+ followed by
Wxd8+.
7...£>f5 (D)
8&xf5
The most common, and also most natural
since it weakens Black's pawn-structure. There
are disadvantages with other ways of defending
the d-pawn:
a) 8 &e3 £>xe3 9 fxe3 £.e7 10 £>c3 0-0 11
0-0 f6 12 exf6 &xf6 13 We2 £.d7 14 2adl
We7 15 Abl £.e8 = Gormally-Arkell, Lichfield
2000.
b) 8 £.c2! ? Wb6 9 Axf5 exf5 10 0-0 &e6 11
^c3 h6 12 £>a4 Wb5 13 £>el &e7 (a possible
improvement could be to postpone this; for
example, 13...2c8 14 .&e3 b6 intending to meet
15 £>d3 with 15...£>b4) 14 &e3 2c8 15 £>d3 b6
(15...£>b4 16£>dc5 is better for White) 16^c3
Wc4 17 2c 1 g5! ? 18 f4, and White was better in
Timman-Andersson, Malmo 2000.
8...exf5 9 £>c3 £.e6 (D)
This is a strategically interesting position.
White has slightly more space if he can maintain
his pawn on e5, and avoid Black gaining space
on the kingside. Black's light-squared bishop is
poor at this moment, but the bishop-pair is a
potential force, and Black naturally hopes the
bishops will have their say later in the game.
10 h4
As in many similar positions in the French,
this type of move is of a more defensive nature
than it is aggressively minded. It is true that
White gains space on the kingside but there is
always a risk of the h-pawn becoming weak

90
FRENCH: ADVANCE AND OTHER LINES
W
when it advances. However, White's primary
aim is to avoid Black gaining space with ...g5.
White's main alternative is 10 0-0 JLe7, and
then:
a) 11 £te2 g5! has been considered fine for
Black since Teschner-Portisch, Monte Carlo
1969.
b) 11 the 1, intending to meet...g5 with f4, is
better but White loses control of the centre, and
in Mesropov-Moskalenko, Moscow 1996 Black
equalized after 1 L..Sc8 12 £ic2 (12 £ie2 0b6
=) 12...0-0 13 0d3 <4>h8 14 Ad2 a6 15 <4>hl
£ia5 16 b3 £ic6 17 h3 £ib4 18 £ixb4 Axb4.
10...h611h5£e7 12£ie2
White will bring his knight to f4, and attempt
to increase the pressure on the kingside with
Sh3-g3. Note that the immediate transfer of the
knight with 10 £te2 (instead of 10 h4) would
have given Black a chance to prevent the
advance of the h-pawn with 10...h6 11 h4 g6!?, as
in Adams-Vaganian, Lucerne Wcht 1997.
12...Va5+ 13 *fl Sc8 14 <&gl <4>d7! (D)
W
Black's king is much safer in the centre, and
this connects the rooks and prepares to double
on the c-file.
15 £if4 fic7 16 Sh3 Shc8 17 Sg3 Af8 18
Ml 0a6 19 Ac3 <4>e8
Khalifman-Akopian, Dortmund 2000. With
patient manoeuvring, Black has achieved a solid
position with good chances of counterplay on
the queenside. Chances are roughly equal.

6 Various Deviations
1 e4 e6 2 d4 d5 3 e5 (D)
ifiif « a W4 a
^
aba'b ha a
We conclude the coverage of the Advance
Variation with a chapter on various early
deviations. These consist mostly of White's 4th
move alternatives (after 3...c5), but there are
also a few alternative ways for Black to play.
The most noteworthy is 3...c5 4 c3 #b6 5 £tf3
JLd7, intending to play ...JLb5.
A: 3...b6 91
B: 3...fce7 92
C: 3...c5 93
Vaganian has played 3...#d7, intending ...b6.
It rules out 3...b6 4 JLb5+, which we see in Line
A, and avoids committing the king's knight to
e7 as in Line B. It has its points, but the theory
of these lines isn't very conclusive. I shall only
put forward the superficial general assessment
that White is slightly better in these types of
positions, but Black remains very solid.
A)
3...b6 (D)
4:ib5+
This is a thematic move in the French against
various ...b6 set-ups. Black's idea is to exchange
W
mp gup p
A if A
BAB BAB
the light-squared bishops with ...JLa6, and
White hopes to prevent this with the text-move.
The idea is that if Black replies ... JLd7, White
will retreat his bishop to d3, and if Black plays
...c6, White goes JLa4, after which he can
complete development with c3, £>e2, etc., when
again he avoids exchanging bishops. White has
two main alternatives:
a) 4 4tf3 Wd7 5 c4 Ab4+ (5...£ie7 6 £ic3
&bl-3...£he74fof3b65c4&.b76£hc3Wd7)
6 £ic3 Ab7 7 cxd5 Axd5 8 Ae2 £ie7 9 0-0
Ab7 10 a3 Axc3 11 bxc3 £ibc6 12 Ab5 a6 13
JLd3 h6 with approximately equal chances,
Zeller-Vaganian, Bundesliga 1995/6.
b) 4 c3 #d7 5 a4 and then:
bl) 5...JLa6 6 JLxa6 £ixa6 and here:
bl 1) 7 a5 is usually considered slightly better
for White but in Sax-Rausis, Cattolica 1994,
Black managed to blockade the position after
7...b5 8 b4 0-0-0 9 £tf3 £>e7 10 £ia3 f6 11 h4
h5 12#d3c6 13£f41/2-1/2.
bl2) In Ye Jiangchuan-Ivanchuk, Moscow
FIDE KO 2001, White tried 7 ftt3 #c8 8 b4
£ie7 9 b5 £ib8 10 £id2 a6 11 Ibl axb5 12 axb5
c6 13 £ie2 cxb5 14 Ixb5 £id7 15 0-0, with a
slight advantage.
b2) 5...a5 6 £id2 £ie7 7 £idf3 c5 8 Ad3
Aa6 9 £ie2 Axd3 10 #xd3 £ibc6 11 0-0 cxd4

92
French: Advance and Other Lines
12 £texd4 £>xd4 13 £)xd4 £)c6 14 £ib5 i.c5
with approximately equal play, Ehlvest-Vag-
anian, Istanbul OL 2000.
We now return to the position after 4 jk.b5+
(D):
IMA.
mi
HI « *
IB
ABA
; mm
AH
WMbWb
mm
4...c6
4...Ad7 5 ±d3 c5 6 c3 £ic6 7 «M3 (we can
compare this with a normal Advance French:
Black has a tempo more in ...b6, but this doesn't
help him much, and White has been able to
develop his bishop on d3, without having to worry
about his d4-pawn) 7...f6 8 0-0 fxe5 9 dxe5?! (9
£>xe5 £>xe5 10 dxe5 ± is better; it is notable
that in many situations similar to this it is
advisable to capture with the knight on e5) 9...#c7
10 Sel £ih6 11 c4 d4, Anand-I.Rogers, Manila
IZ 1990, and now White's best chance of an
advantage is 12 ±xh6 gxh6 13 £>bd2 ±.
5 ±a4 b5 6 ±b3 c5 7 c3
7dxc5!?.
7...fce7 8 4M3
8 £>e2, to leave a path open for the f-pawn,
could also be considered.
8...£iec6 9 0-0 h6!? 10 ±e3 £id7 11 £ibd2
c4 12 ±c2 b4
Black has good counterplay on the queen-
side, Shabalov-Seirawan, USA Ch (Chandler)
1997.
B)
3...fce7 (D)
Black wants to follow this with a ...b6 set-up
but ...4&e7 is a useful preparatory move, and
W
* A * SiBl
"""^ /''"'/'wd''±WB. '""/'WM.
m mm* mm
I ll ■ !m
in in ii
wmm&m
Black hopes that White's next move will make
the .&b5+ ideas less attractive.
4£if3
4 c3 is maybe the most cunning waiting
move, when 4...b6 can be met by 5 .&b5+, but
Black can also react with a clever answer:
4...#d7, again preparing ...b6.
4...b6 5 c4 (D)
This is the most direct approach. White
hopes to benefit from his slight lead in
development and space advantage, but the idea is also
slightly anti-positional because Black has good
chances of getting control of d5. Another
option is 5 c3 Wd7 6 £ibd2 a5 (6...Aa6 7 ±xa6
£ixa6 8 a4 ±) 7 h4!? h5 (I don't really like this,
and would prefer 7...JLa6 with roughly equal
chances; White will get some space on the
kingside but Black obtains counterplay on the
queenside with ...c5 and ...£>ec6) 8 £>g5 .&a6 9
±xa6 £ixa6 10 «tfl 4tf5 11 £>g3!? (11 £ie3
looks more sensible) ll...£>xg3 12 fxg3 f6 13
exf6 gxf6 14 £ih3 0-0-0 15 0-0 e5!? 16 Sxf6
±g7 17 Sf3 exd4 18 cxd4 Sde8 with
counterplay, Kupreichik-Vaganian, USSR 1984.
5...±b7 6 £ic3 Wd7 7 cxd5
7 ±e2 £ibc6 8 0-0 dxc4 9 ±xc4, Sax-Short,
London 1980, and now 9...0-0-0! 10 ±e3 £tf5
11 d5 £ixe3 12 fxe3 exd5 13 #xd5 #xd5 14
.&xd5 .&c5 looks fine for Black.
7~®xd5 8 ±d3 c5 9 0-0 £ixc3 10 bxc3
cxd4
Kupreichik mentions 10...£>c6 11 dxc5,
giving ll...l.xc5 12 ffe2 Sd8 13 Ax2 as slightly
better for White, but 11...0-0-0! is a clever
finesse, securing Black a pleasant position.

Various Deviations
93
11 £ixd4
11 cxd4 £ic6 12 Ae4 Sd8 13 Ae3 £ie7 is
equal according to Kupreichik.
Il...£ic6 12 Sbl Ac5 13 Ae3
Kupreichik-Vaganian, USSR 1980. Now the
simplest equalizer is 13... JLxd4 14 cxd4 £>e7.
C)
3...C5 (D)
W
\,4
/s^tf/" m //i'^i//'
tomb JL
k m #IA|
HAS
■ ABA
I «
^
AHAI
tJAH
Now:
CI: 4#g4
C2: 4dxc5
C3: 4£>f3
C4: 4c3
93
94
95
96
White has tried a number of more obscure
moves. I shall restrict myself to mentioning 4 f4,
which looks like a fairly natural move. If Black
plays 4...cxd4, White can, by 5 £tf3 £>c6 6 Ad3,
try a gambit in the spirit of 4 tfcfi £hc6 5 Ad3
cxd4. However, the pawn is not particularly
helpful for White on f4 and in the broader sense
only restricts White's dark-squared bishop.
Andreev-Alexeev, Ukrainian Ch (Simferopol)
2003 is a warning example: 4...£>c6 5 £rf3 £>h6
6 Ad3 cxd4 7 a3 ±d7 8 0-0 &e7 9 b4 a6 10
We2 fb6 11 &b2 £>g4 12 4>hl 0-0-0 13 £sbd2
f6 14 £ib3 h5 15 £ibxd4 fxe5 16 fxe5 Sdf8 17
£>xc6 &xc6 18 h3 We3 19 hxg4?? hxg4+ 20
£>h2 2xh2+0-l.
CI)
4 #g4 (D)
v///m vim Lm
'm.
i IP 4 up 4
I 11*11!*
if a «
^ Ml * iJ^
I B *
mi
o&m
*77^/l-7J %^
This was played a few times by Nimzowitsch
but is probably a little too sharp for his ideas of
over-protection. Indeed, White is mainly
interested in keeping his pawn on e5, and is ready to
sacrifice his d-pawn for it. The early queen
move is intended to make Black's development
of his kingside more problematic but, not
surprisingly, also gives Black plenty of tactical
resources. Sveshnikov, a great expert in the
Advance Variation, played it only a few years ago,
which suggests that it should be taken just a
little seriously.
4...cxd4
4...^a5+ has been considered a safe and less
demanding alternative. Neither interposition by
a piece is without obvious drawbacks. I have my
doubts about 5 c3 cxd4 6 £rf3 dxc3 (6...^c6
seems a sound alternative) 7 £>xc3, but this
could be White's best, and indeed has scored

94
French: Advance and Other Lines
100% in the two games with it that I have seen.
White is clearly ahead in development and
Black still needs to find a way to develop his
kingside.
5 £>f3 £>c6 6 Ad3 #c7 (D)
The most direct. 6...&ge7 7 0-0 £>g6 8 lei
(8 #h5!?) 8...&e7 9 c3!? dxc3 10 £>xc3 Ad7
11 Ad2 £>b4 12 Abl Ic8 13 a3 £>c6 14 Ad3
0-0 15 h4 f5 16 exf6 Ixf6 17 Ag5 Ixf3 18
#xf3 Axg5 19 Axg6 ±xh4 20 Af7+ <S?h8 21
Sadl was better for White in Sveshnikov-
Komarov, Vrnjacka Banja 1999.
W
9&§
7#g3
Or:
a) 7 &f4 £>ge7 8 Ag3 £>b4!? 9 0-0 £>xd3
10 cxd3 ^f5 was good for Black in Ruzicka-
Vavra, Prague 1992.
b) 7 0-0 £>xe5 8 £>xe5 #xe5 9 &f4 £>f6! 10
#xg7?! (this is not good, but 10 *Tg3 #h5 is
also very good for Black) 10...#xf4 11 #xh8
^>e7! and Black wins - Remlinger. The threat is
12...#h6andl3...1.g7.
7...f6! 8 exf6 #xg3 9 f7+ <S?xf7 10 hxg3 e5
Black is slightly better - Fine.
C2)
4 dxc5(D)
In the Advance Variation, White's prime
concern is to maintain a strong centre. With this in
mind, White pays particular attention to his
e5-pawn - if it can be maintained, then it
guarantees a nice space advantage on the kingside.
Here, White intends to guard his e-pawn with
Ml H
i
#fAfi n
i §
%%^ mm, w//a
B B H ■ ,
ft if ft 9 if ft iH
w.
h&bothaei
his pieces but Black obtains excellent counter-
play if he attacks White's centre with ...f6 at the
right moment.
4...£>c6
This is the most accurate. On 4...JLxc5
White would play 5 #g4.
5 £>f3 ±xc5 6 Ad3 £>ge7 (D)
The most common. Black wants to finish
development before embarking on a break against
the centre. The other option, 6...f6 7 #e2 fxe5 8
£>xe5, should also be fine for Black. Then:
a) 8...£rf6 9 Af4 0-0 10 0-0 £>e4 11 £>xc6
bxc6 12 Ae3 Axe3 13 #xe3 £>f6 14 £>d2,
Becker-Maroczy, Karlsbad 1929, and now
Becker gives 14...#b6!=.
b) 8...£>xe5 9 #xe5 £>f6 (9...#f6!?) 10
Ab5+ *f7 11 0-0 ±d6 12 tfe2 If8 13 f4 *g8
with roughly equal chances, Feoktistov-Yantu-
rin, Moscow 1998.
W
aBa«"WaS
7 AM

Various Deviations
95
The alternative is 7 0-0. When looking at this
line on my database I was wondering why this
position was so popular at grandmaster level,
until I realized that it only arises frequently
with White having to make the next move too!
The explanation is that the position arises from
a Caro-Kann: I e4c6 2 d4 d5 3 e5 c5 4 dxc5 e6
5 %sf3 ±xc5 6 &d3 £hc6 7 0-0 *hge7. Thus,
compared to the Caro-Kann line, where Black
incidentally scores very well, Black has an
extra tempo and can prevent the natural
development of White's dark-squared bishop: 7...£>g6
8 Sel #c7 9 tfe2 £id4 10 £ixd4 Axd4 11
JLxg6 hxg6 12 c3 JLb6 = Bebchuk-Belinkov,
USSR 1967.
7...#b6 8 0-0 £ig6 9 Ag3 #xb2 10 £ibd2
£>gxe5 11 £ixe5 £ixe5 12 Sbl #c3 13 Sb3
#d4
White does not have enough for the pawns,
Th.Johansson-E.Berg, Bergen 2001.
C3)
4GX3(D)
An idea similar to 4 #g4. White is still ready
to sacrifice his d-pawn, hoping that it will only
be temporarily, in return for quick development
and a strongly supported e-pawn.
4...£ic6
This is Black's most logical move, although
there are also other ideas, delaying the
development of this knight. After 4...cxd4 White
normally continues development rather than
recapturing the pawn:
a) 5 £>xd4 £>c6 gives Black easy equality;
e.g., 6 £ixc6 bxc6 7 Ad3 £ie7 8 0-0 £ig6 9 f4
Ae7 10 c3 #b6+ 11 *hl c5 = M.Poulsen-
Bern, Internet 2000.
b) 5 #xd4 £ic6 6 Wf4 (6 Wg4?! #c7 ?)
6...Wc7 7 £ic3 a6 8 Ad3 f6!? 9 exf6 #xf4 10
Axf4 gxf6 11 0-0 £ige7 12 £ia4 e5 13 £ib6
Sb8 14 JLg3 JLe6 and Black is slightly better
due to his strong centre, Rumens-Short, British
Ch(Ayr) 1978.
c) 5 JLd3 (D) and now:
cl) 5...&c6-4...*hc6 5 &d3 cxd4.
c2) 5...*fb6 6 0-0£id7!?7 £ibd2 (gambiting
a pawn with 7 c3 deserves attention) 7...£>e7 8
£ib3 £ic6 9 Sel g6 10 Af4 Ag7 11 #d2 0-0 12
h4, Keres-Stahlberg, Warsaw OL 1935, and
now Kosten suggests 12...a5 13 a4 £>c5 but
gives no clear assessment. Presumably he thinks
Black is OK. I also quite like 12...f6!?.
c3) 5...£te7 6 0-0 £ig6 (Black's idea of
delaying the development of his queen's knight
has prevented White from playing JLf4) 7 Sel
£ic6 8 £ibd2 (8 g3 #c7 9 tfe2 Ae7 10 h4 f6 11
Axg6+ hxg6 12 Af4 #b6! 13 Wd3 *f7 14
JLcl JLd7 15 a3 Sae8 is very good for Black,
Skripchenko-Matveeva, Belgrade worn 2000)
8...Wc7 9 Axg6 hxg6 10 £ib3 f6 (10...&d7 is a
safe alternative) 11 £ibxd4 £ixe5 12 Af4 (12
£ixe5 fxe5 13 £ib5 #b6 14 c4 a6 15 £ic3 1x5
16 Se2 oo) 12...£ixf3+ 13 #xf3 #b6 14 #xd5
*f7 15 #c4 Ac5 16 Sadl ± Beikert-Schlind-
wein, 2nd Bundesliga 1997/8.
5 ±63 cxd4 6 0-0 (D)
6...£ige7

96
French: Advance and Other Lines
A 2*^ *£?
I iisU
Other moves:
a) 6...£c5 7 Af4 £>ge7 8 £ibd2 £ig6 9
Ag3 Ad7 10 £ib3 Ab6 11 Sel Sc8 12 h4 ±
Spraggett-I.Ivanov, Canada 1986.
b) 6...f6!? 7 We2 Wc7 (7...fxe5 8 £>xe5 £ixe5
9 Wxe5 4M6 10 Af4 Wb6 11 £id2 Ae7 12 £ib3
0-0 13 Ag3 h6 14 Wc7 Wxc7 15 Axc7 Ad7 16
Sfel ± Van Wijgerden-Van der Sterren, Dutch
Ch (Leeuwarden) 1977) 8 Ab5 (8 Af4 g5 9
Ag3 g4 10 £ih4 and instead of 10...f5 11 f3
£ih6 <*> Bryson-Zuger, Manila OL 1992,
Watson suggests 10...fxe5!, with a massive pawn-
centre) 8...fxe5 9 £ixe5 Ad6 10 Wh5+ g6 11
£ixg6 Wf7 (11...4M6 12 Wh4 Axh2+ 13 *hl
Sg8 14 Wxf6 hxg6 15 g3 Wh7 16 *g2 Sh8 17
Af4! is not clear at all) 12 £tf4 £>f6 =.
7 Af4 £ig6 8 Ag3 Wb6 9 £ibd2 (D)
bishop-pair. Instead, the greedy 9...Wxb2 10
£ib3 (10 Sbl!? Wxa2 11 £ib3 also gives White
compensation) 10...£e7 11 h4 0-0 12 h5 £ih8
13 Af4 Wa3 14 h6 g6 15 £ibxd4 £ixd4 16
£ixd4 Wb4 17 Ae3 f6 18 Sbl Wa4 19 exf6
JLxf6 20 £>b5 gave White compensation in
Bator-Bareev, Stockholm 1987.
10 £ib3 £ixd3 11 Wxd3 £&7 12 £ifxd4
Ae713f4 0-0
with approximately equal play, Seul-M.Gure-
vich, Bad Godesberg 1996. White has good
control of d4 but without his light-squared
bishop much of the dynamism and attacking
potential of the white position is lost.
C4)
4 c3 (D)
m
t^fa^m
IH±H*H*
^M^
Mi
w/' A fay///
W,
AH ■ BAP
9...£ib4!?
This gives up the hope of keeping the extra
pawn but it is almost always useful to take the
Here we discuss:
C41: 4..J?b6 96
C42: 4...£ic6 100
In C42 we focus on Kupreichik's 5 JLe3, as
the other significant lines have been covered in
previous chapters.
C41)
4...WM 5 4M3 Ad7 (D)
By delaying the development of the b8-knight
Black now seeks to exchange light-squared
bishops with ... JLb5. Obviously, Black can also
play 5...£>c6, thus transposing to 4...^hc6 5 ^hf3
Wb6. This move-order is frequently chosen if
Black already knows that he will be playing the

Various Deviations
97
5...Wb6 line but wants to avoid Kupreichik's 5
±e3 (Line C42).
6±e2
Or:
a) 6 £ia3 cxd4 7 cxd4 ±b4+ 8 ±d2 £ie7 9
i.xb4 #xb4+ 10 #d2 #xd2+ 11 <A>xd2 a6 12
i.d3 £ibc6 13 £ic2 f6 = A.Kiselev-Eshchenko,
Kiev Ch 2000.
b) 6 £ibd2 cxd4 7 £ixd4!? £ic6 8 £>2f3
£>xd4 9 £ixd4 £ie7 10 ±d3 £ic6 11 £tf3 d4! 12
0-0dxc3 13bxc3±e7 14#e2h6 15a4#c7 16
#e4Sc8 17±e3£ia5 18Sabl b6 19#g4<A>f8
with equality, Dvoirys-De Jong, Hoogeveen
2000.
c) 6 ±d3 (D) and now:
cl) 6...±b5 7 dxc5 Axc5 8 b4 Af8 9 0-0
±xd3 10 #xd3 £ie7 11 £ia3 £ibc6 (ll...a6?!
12 Sdl £id7?? 13 £ic4! 1-0 Jonkman-Afek,
Wijk aan Zee 1997) 12 ±e3 #d8 13 ±g5 Wb8
14 £ib5 £ig6 15 Sael h6 16 Acl ±e7 ?
Bokan-Kinsman, Cannes 1998.
c2) 6...cxd4 7 £ixd4 (7 cxd4 i.b5 =) 7...£ic6
(7...£>e7!? could be more accurate) 8 £>xc6
i.xc6 9 #e2 £ie7 10 £id2 £ig6 11 £>f3 i.e7 12
0-0 #c7 13 ±d2 ±d7 14 Sael 0-0 15 £id4
±c5 16 #h5 ±xd4 17 cxd4 f5! (Black must
react to the threat of White swinging a rook to the
h-file) 18 exf6 Sxf6 19 ±g5 Sf7 20 ±xg6
hxg6 21 #xg6 #b6 22 Sdl (22 Se3 #xd4 23
Sh3 Sf5 24 #h7+ *f7 25 #h5+ g6 26 #h7+
#g7 with the idea 27 f4 d4! looks OK for
Black) 22...#xb2 23 Sd2 #c3 24 h4 Saf8 25
Sbl b6 = Grishchuk-M.Gurevich, Wijk aan
Zee 2002.
d) 6a3!?(T>)andthen:
dl) 6...£hc6-4...£hc65£hf3Wb66a3&d7.
d2) 6...cxd4 7 cxd4 ±b5 8 ±xb5+ #xb5 9
£ic3 #a6 10 £ie2 £ie7 11 0-0 £id7 12 £ig3 £ic6
13 .&e3 and now 13...h6 was approximately
equal in Alekseev-Guseinov, Artek 1999 but
Black could also consider not weakening his
kingside and simply continue 13...^.e7 and
...0-0.
d3) 6....&b5 and now:
d31) 7 b4 cxd4 8 ±xb5+ #xb5 9 cxd4 (9
£ixd4 #d7 10 0-0 £ie7 11 Af4 £ibc6 12 4tf3
h6 13 h4 g5 14 hxg5 £ig6 15 ±g3 hxg5 16
£>xg5 £>cxe5, Kupreichik-Feigin, 2nd Bundes-
liga 1999/00, and after 17 f4 ±h6! the
complications should favour Black) 9...£>d7 (9...a5 10
£ic3 #c6 11 ±d2 axb4 12 axb4 Sxal 13 #xal
* Zviagintsev-Volkov, Russia Cup (Samara)
1998) 10 £ic3 #c6 11 ±b2 £ib6 12 £id2 £ie7

98
French: Advance and Other Lines
13 0-0 £rf5 14 2c 1 #d7 = Alekseev-Ruste-
mov, Russian Cht (Tomsk) 2001.
d32) 7 c4 (D) and here:
W§ WB iff A M
d321) 7...Axc4 8 ±xc4 dxc4 9 £ibd2! #a6
10 tfe2 £>e7 (10...cxd4 11 £>xd4 ±c5 12 £>4f3
c3 13 bxc3 #xe2+ 14 4>xe2 £>d7 15 £>e4 ±
Sveshnikov) 11 £>xc4 £>d5 12 0-0 £>c6 13 dxc5
±xc5 14 b4 b5 (14...±xb4 15 axb4 #xal 16 b5
£>d4 17 ^d6-h 4>d7 18 £>xd4 #xd4 19 £>xb7 is
unclear) 15 £>d6+ ±xd6 16 exd6 2d8 17 ±b2
± Sveshnikov-Matveeva, Cappelle la Grande
1995.
d322) 7...dxc4 8 £>c3 cxd4 (8...£>c6 9 d5
0-0-0 10 d6 f6 11 £>xb5 #xb5 12 #c2 #b3 13
#xc4 #xc4 14 ±xc4 £>xe5 15 £>xe5 fxe5 16
±xe6+ ^>b8 17 d7 ± Fressinet-Sadvakasov,
Calicut jr Wch 1998) 9 £>xb5 #xb5 10 #xd4
£>c6 11 Wf4 (11 #xc4 #xc4 12 ±xc4 ±c5 13
b4 ±b6 14 ±b2 £>ge7 15 0-0 0-0 16 Hfdl 2fd8
17 ±d3 h6 18 ±e4 £>d5 19 g3 was minimally
better for White in Svidler-Gelfand, Moscow
FIDE KO 2001) ll...*fa5+ 12 ±d2 ±b4 13
±xc4 ±xd2+ 14 #xd2 #c5 15 #c3 £>ge7 16
b4 #b6 17 0-0 0-0 18 ladl 2ad8 19 ±d3 ±
Svidler-Gelfand, Moscow FIDE KO 2001.
We now return to 6 ±e2 (D):
6...cxd4
This has taken over as Black's main choice.
6...£>c6 transposes to 4...^hc6 5 £hf3 !Ld7 6
§Le2 %b6, while the 'older' 6.. JLb5 can lead to
sharp positions:
a) 7 dxc5 ±xc5 8 0-0 ±xe2 9 #xe2 a5! ? 10
a4 (10 £>bd2 a4 =) 10...^e7 11 £>a3 0-0 12
£>b5 £>d7! (the knight is better placed here,
IA
I
m m
w^.
W, * Iff f
«
leaving c6 free for the other knight, thus
exerting maximum pressure against e5) 13 .&d2
£>c6 = Lobzhanidze-Volkov, Turheim 1999.
b) 7c4!?(T>).
This is the main reason that Black has begun
to exchange on d4 before playing ....&b5, but it
is still far from resolved:
bl) 7...dxc4 deserves more attention. It is
interesting that this idea has recently received
attention in the line 6 a3 Jk.b5 7 c4. Now:
Ml) 8 £>c3 cxd4 and then:
bill) 9#xd4±c6 10#xb6axb6 11 ±xc4
£>d7=.
bll2) 9 £>xb5 ±b4+!? 10 *fl #xb5 11
#xd4 £>c6 12 #xc4 #a5 13 a3 ±.
bl 13) 9 £>xd4 ±a6!? (9...±c6 has also done
well for Black) 10 ±e3 #a5 (10...#xb2 112c 1
oo) 11 f4 £>h6 12 #c2 ±c5 13 0-0 £>d7 14 £>e4
with compensation for White, Yagupov-Ruste-
mov, Russian Cht (Moscow) 1994.

Various Deviations
99
bl2) 8 d5 exd5 9 #xd5 £ie7 10 tfe4 %6
(10...#c6 =) 11 #xg6 £ixg6 12 £ia3 Aa6 13
JLxc4 JLxc4 14 £>xc4 £>c6 = Galdunts-Gra-
barczyk, Gemany tt 1999/00.
b2) 7...£xc4 8 Axc4 #b4+! (8...dxc4 9 d5
£ie7 10 dxe6 fxe6 11 0-0 4tf5 12 £ia3 Ae7 13
£ixc4 #a6 14 b3 0-0 15 Ag5 £ic6 16 ttt7 £ib8
17 Wd3 £ic6 18 Axe7 £icxe7 19 Sadl ± Benja-
min-Nogueiras, Lucerne Wcht 1997) 9 £bd2
dxc4 (D) and now:
9 £ixe2 (T>)
W
b21) 10 0-0 cxd4 (10...£ic6 11 dxc5 Axc5
12 a3 Wb5 13 £ie4 £ige7 14 Ae3! Sd8 15 tfe2
Axe3 16 £id6+ Sxd6 17 exd6 Axf2+ 18 Sxf2
£\f5 19 fid 1 and White was better in Adams-
Illescas, Las Palmas 1994) 11 £ixd4 £id7! 12
tfe2 fic8 13 £>4f3 £ie7 14 a3 #a4 15 £ie4 £ig6
16 Ag5 h6 17 Ad2 £ic5 18 £ixc5 Axc5 +
Sandor-Mann, Bundesliga 1995/6.
b22) 10 a3 (this move intends to improve
the move-order by discouraging the ...cxd4
option) 10...#b5 11 0-0 £ie7! (Il...£ic6 12 dxc5
Axc5 -10 0-0 £hc6 11 dxc5 &xc5 12 a3 1*5)
12 £ie4 (12 dxc5 £id5 13 tfe2 Axc5 14 £ixc4
might be a better try) 12...£>d5 13 dxc5 £>d7 14
£>d6+ JLxd6 15 cxd6, Illescas-Beliavsky,
Madrid 1995, and now Illescas gives 15...c3! as at
least equal for Black.
7 cxd4 Ab5 8 £ic3
8 0-0 Axe2 9 Wxe2 #a6! is considered equal.
8...Axe2
8...AM 9 0-0 Axc3 10 Axb5+ #xb5 11
bxc3 £ie7 12 a4 Wd7 13 Aa3 £ibc6 14 Wd3
fic8 15 a5 gave White an edge in Degraeve-
Vaisser, French Cht 1998.
9...£>e7!
Black's main concern is how to place his
knights. Here, it seems best to place them on c6
and d7, from where they can most easily be
used in counterplay on the queenside. Other
moves:
a) 9...£>c6?! (this natural move is probably
wrong) 10 0-0 £ige7 11 £ig3 fic8 (1 l...£ig6 12
£>h5 is awkward for Black as he cannot develop
his bishop) 12 Ae3 g6 13 Ag5 Ag7 14 #d2
and White was better in Nunn-J.Polgar,
Monaco Amber blindfold 1994.
b) On 9...&b4+ Psakhis gives 10 <4>fl!? as
White's best.
10 0-0 ^totfllfiblaSf/))
w
12 mil?
Grishchuk's improvement on one of his own
games. The queen obstructs the bishop, but

100
FRENCH: ADVANCE AND OTHER LINES
White's idea is simply to transfer the queen to
the kingside. In Grishchuk-M.Gurevich, Cannes
rpd 2001, White obtained little from 12 &e3
£>d7 13 h4 £e7 14 £g5 h6 15 &xe7 £>xe7 16
Wd2 0-0 =.
12...H613 h4 £>d714 Sdl Wa615 a3 a416
£>c3£>a517Wf4
White is slightly better, Grishchuk-Vagan-
ian, Erevan Wcht 2001.
C42)
4...£>c6 5 ±e3 (D)
EV/////// A. '&19& + Y/zMr/A M^
■J.lt*Hft
VA0 ilii
■ Hill ^
Kupreichik's idea. The reason behind it is
that White often has difficulties developing the
bishop in the 5 £tf3 Wb6 line. White can now
meet ...Wb6 with Wd2, and also stays flexible
in respect to how to arrange his kingside,
leaving open the possibility of playing f4.
5...Wb6
5...£>h6 6 £tf3 £>f5 7 ±d3 cxd4 8 cxd4 - 5
Qtf3 *hh6 6 k.d3 cxd4 7 cxd4 £>/5 8 k.e3.
6 Wd2 Ml 7 £>f3 cxd4
This exchange gives White the option of
playing his knight out on c3 but it still seems
advisable in order to keep the centre stable.
Other moves:
a) 7...Sc8 8 dxc5! £xc5 9 £xc5 Wxc5 10
Wg5 £>ge7!? (10...Wf8 11 ±d3 f6 12 Wg3
£>ge7 13 £>bd2 Wf7 could also be considered)
11 £>bd2 (11 Wxg7 Sg8 12 Wxh7 d4! gives
Black a lot of counterplay) 11...0-0 12 £d3
Wb6!? 13 JLxh7+ (this secures a draw but no
more; 13 £lb3 is a sensible alternative but
Black can then counter with 13...f6!?, which is
what Black's last move prepared, whereas the
immediate 12...f6? would have been a blunder
in view of 13 exf6 Sxf6 14 £>e4! +-) 13...<i>xh7
14 Wh5+ 4>g8 15 £>g5 Sfe8 16 Wxf7+ 4>h8 17
Wh5+ <A>g8 18 £>df3 (White can take a draw
whenever he wants, but first decides to test
Black's defensive skills; it is not easy at all for
Black to find useful moves) \%..Mb5 (18...Wxb2
19 0-0 Wxc3 20 Sael Wd3 also defends) 19
0-0-0 Wc4! 20 g3 Wxa2 21 Wf7+ <&h% 22
Wh5+ <A>g8 23 Wf7+ &\& V2-V2 Kupreichik-
Dolmatov, Yugoslavia 1992.
b) 7...£>h6 8 £e2 (8 ±xh6 gxh6 is generally
not a problem in these types of positions; 8 &.(&
c4 9 £e2 £>f5 =) 8...£>g4 9 0-0 £>xe3 10 fxe3
£e7 11 £>a3 f6 12 exf6 ±xf6 13 <&h\ 0-0 was
fine for Black in Kupreichik-Visser, Groningen
1996.
8cxd4 2c8 9£>c3£>a5
It is a typical scenario in the Advance
Variation that Black seeks counterplay on the queen-
side before completing development of his
kingside.
10 ±d3 ±b4 11 0-0 (D)
B
. mmm
iiiiii
A
AH
ll...£>c4
ll...£>e7 12 Sabl!? £>c4 13 We2 Wa5 14
Sfcl h6 (Kupreichik suggests 14...^xe3 15 fxe3
h6, when, for example, 16 a3 JLxc3 17 Sxc3
Sxc3 18 bxc3 b6 looks equal) 15 a3 £xc3 16
Sxc3 0-0 (16...£>xe3 17 2xc8+ ±xc8 18 Wxe3
£d7 19 Scl ±) 17 ±xh6?!! (this speculative
sacrifice is typical of Kupreichik's style:
probably not correct, but very dangerous) 17...gxh6

Various Deviations
101
18 £}g5! (the idea of White's previous move,
but it is far from clear that White's sacrifice is
correct) and now the fine attacking game Ku-
preichik-Kaminski, Bad Endbach 1995
continued 18...£>xb2? 19 ±h7+! &hS (19...*g7 20
Sh3! hxg5 21 Wh5 f6 22 Wh6+ *f7 23 exf6
4>e8 24 Wxg5! Sxf6 25 Wxf6 Wd2 26 ±c2!
&dS 27 Sh7 and White wins - Kupreichik) 20
Sh3 ±b5 21 Wf3 f6 22 2xh6, and White
crashed through. The critical line is 18...hxg5!
19Wh5f5!(Dj:
a) 20 Wxg5+ *f7 21 Wh5+ 4>g7! 22 ±xc4
^g6! 23 Sg3 ±e8 and White's attack comes to
a dead end.
b) 20 ±e2? ±e8! 21 Wxg5+ £>g6 with the
ideaof...Wd8-+.
c) 20 exf6 Sxf6 21 #xg5+ *f7 22 #h5+ is
White's best chance but 22...<4>g7 23 ±xc4!?
(23 Wg5+ £>g6 24 ±xg6 Sxg6 25 We7+ 4>g8
26 Wxd7 Wc7 +; 23 Wh7+ *f 8 24 Wh8+ £>g8!
+) 23...£>g6 doesn't look convincing for White.
12 ±xc4
12 We2 should still be considered, in
comparison with the Kupreichik game of the last
note. Black can hardly have anything better
than 12...£te7, as 12...£>xe3 13 Wxe3!? (13
fxe3 ±xc3 14 bxc3 Sxc3 15 £>g5 f5 16 exf6
£>xf6 oo) 13...Axc3 14 bxc3 Sxc3 15 Wg5 could
be dangerous.
12...Sxc413Sfcl
13 a3 JLxc3 14 bxc3 Wc7 is equal, according
to Korchnoi.
13...£>e7 14 Wdl ±xc3 15 bxc3 0-0
= Trygstad-Hanley, Gausdal 2001.

7 Exchange Variation
1 e4 e6 2 d4 d5 3 exd5 exd5 (D)
The Exchange Variation does not trouble
Black from a theoretical viewpoint, but
psychologically many players find it difficult to
face. Much of the life is immediately taken out
of the position and it is far from easy to do
anything to change the symmetrical structure that
arises. This is not really a situation that suits
many French players, but in practice the game
often livens up, unless White is very
unambitious. In fact, the Exchange French is very
similar to the Petroff, which has been
sharpened greatly in many lines in recent decades.
Structurally the two openings are identical but
White's chances of an opening advantage are
worse in the Exchange French, since Black's
knight is safer back on f6, compared to e4 in
the Petroff.
We shall look at:
A: 4£>c3 102
B: 4±d3 103
C: 4c4 104
D: 4£>f3 106
There are also a number of other moves, but
listing them all makes little sense, as Black
easily equalizes. Just two short examples:
a) 4 &f4 ±d6 5 ±xd6 #xd6 is a sequence
that is not uncommon but just equal.
b) 4 #f3 £}c6 (one of several options for
Black) 5 £b5 £tf6 6 £>e2 £g4 7 #c3 ±d6! 8
±xc6+ bxc6 9 #xc6+ £d7 10 #a6 £>g4!?
(setting up a little trap that White walks into;
10...0-0 is also simple and good) 11 h3? #f6 12
£>bc3 (after 12 hxg4 ±b4+ Black wins the
queen) 12...*xf2+ 13 <4>dl Wxgl and Black
wins, Av.Bykhovsky-Ulybin, Aaland/Stock-
holm 1997.
A)
4 £>c3 (D)
4...£>f6
4... JLb4 transposes to a Winawer line, 3 £ki
A.b4 4 exd5 exd5, which was examined in my
book The Main Line French: 3 £ki, so here I
shall restrict myself to the text-move.
5±g5
Or:
a) 5 ±d3 £>c6 6 £>ge2 £>b4 7 Af4 c6 8 0-0
£e7 9 h3 0-0 10 £>g3 £>xd3 11 #xd3 £>e8 12
Sfel £>d6 13 Se2 ±e6 14 Sael #d7 15 £>dl
Sfe8 16 c3 Af8 = Hennings-Knaak, Bundes-
liga 1990/1.

Exchange Variation
b) 5 MA Ml 6 Ad3 0-0 7 Wf3 £>c6 8
£>ge2 £>b4 9 h3 £>xd3+ 10 #xd3 £>h5 11 Ae5
g6 12 0-0-0 c6 13 *bl b5 with counterplay,
Landa-Morozevich, Tomsk 1998.
5...M1 6 &d3 (D)
White's most common set-up involves
playing Jtd3 and £}ge2. This can be a strong
attacking set-up as White's knight can move on to f4
or g3, thus getting closer to Black's king, and
avoid being pinned by ...£igA (which is a
common reply if the knight goes to f3 instead).
6...^c6 7£>ge2£>b4
Just so White doesn't get time to keep his
bishop with a3. Another option is to bring out
the bishop with L.AgA 8 f3 Ah5 but this
would be better for White after simply 9 0-0.
White would then have chances of actually
keeping his light-squared bishop, and Black
might run into some trouble with his d-pawn if
he isn't careful.
8£tg3!?fDj
8 0-0 0-0 9 £>g3 £>xd3 10 #xd3 h6 =.
8...£>e4
An interesting tactical solution. 8...£>xd3+ 9
#xd3 c6 10 0-0-0 gave White an edge in Hoi-
berg-A.Matthiesen, Arhus 1998, and if 8...0-0
White might consider keeping the bishop with
9 Ml.
9 Mel
This is the only attempt to play for an
advantage. 9 McA .&xg5! might even be better for
Black.
9...£>xc3
9...#xe7? 10 ±xe4 dxe4 11 a3 ±.
10 &xd8 £>xdl 11 Ixdl *xd812 c3 £>xd3+
13 Ixd3 a5
= Short-Morozevich, Sarajevo 2000.
B)
4 &d3 (D)
This and Line D (4 £tf3) often lead to
symmetrical positions. Here we shall look mainly at
set-ups where White develops his knight to e2.
4...i.d6
If Black is looking for asymmetrical piece-
play, he often chooses 4...£>c6 (4...c5 is also
good) 5 c3 (5 £tf3 -4&J3 *hc6 5 kd3) 5.. ±6.6,
and now:
a) 6 £>e2 #h4!? 7 £>d2 Ag4 8 £>fl £>f6 9
£>e3 0-0 10 h3 2ae8 11 g3 #h5 12 £>xg4 £>xg4
13 *fl «tf6 14 *g2 £>d8 15 £tf4 V2-V2 Blatny-
Kaidanov, Stratton Mountain 1999.
b) 6 Wf3 £>f6 and then:

104
French: Advance and Other Lines
bl) 7 £g5 £g4 8 ®xg4 (8 tfe3+ *d7 °°)
8...£>xg4 9 ±xd8 £>xd8 10 h3 £>f6 11 £>f3 =
West-Smerdon, Canberra 2001.
b2) 7 h3 0-0 8 £>e2 Ie8 9 ±g5 ±e7 10 ±e3
£>e4! 11 ±xe4 (11 £>d2 = is safer) 1 l...dxe4 12
#xe4 £>b4! 13 'idl £>d5 14 £>f4 £>xe3+ 15
fxe3 JLd6 with compensation, Malaniuk-Psa-
khis, USSR Ch (Moscow) 1983.
c) 6 £>f3 ±g4 7 0-0 £>ge7 8 lei #d7 9
£lbd2 (D) and now Black should have good
play whichever side he castles - the choice
probably depends on one's mood:
cl) 9...0-0 10 £>fl Af5 11 £>g3 ±xd3 12
#xd3 ±xg3 13 hxg3 Iae8 14 b3 £>c8 = Slo-
bodjan-Luther, Schoeneck 1996.
c2) 9...0-0-010b4^g611b5^ce712#a4
4>b8 13 £>e5 ±xe5 14 dxe5 ±f5 15 ±f 1 £>c8 =.
Thesing-Bunzmann, Bundesliga 1998/9.
We now return to the position after 4... JLd6
(D):
5#f3
Or:
a) 5Qtf3-4Qsf3&d6 5&d3.
b) 5 £>e2 £>e7 6 0-0 0-0 7 Af4 Af5 only
leads to many exchanges and a dead equal
position; e.g., 8 ±xd6 #xd6 9 £>bc3 c6 10 Axf5
£>xf5 11 ttt3 #f6 12 Ifel £>d7 13 £>g3 £>xg3
14 hxg3 2fe8 V2-V2 Magomedov-Av.Bykhov-
sky, Kurgan 1993.
c) 5^3c6 6&ge2Wh4!?(6...Wf6 7Ae3
Af5 8 Axf5 #xf5 9 #d2 £>f6 = Volkov-
Malaniuk, Smolensk 1997) 7 £e3 £>h6 8 #d2
£ig4 9 g3 #f6 = Minibock-Beliavsky, Vienna
1986.
5...c6
By defending the d-pawn Black is now
planning ...#f6 to exchange queens. The
symmetrical position is then without any prospects for
either side. Instead, if Black is more ambitious
he should choose 5...^e7 but also here it is
difficult to play for more than just a level position.
However, in Barua-Rozentalis, Antwerp 1993,
Black managed something: 6 c3 (not yet
necessary) 6...&d7 7 £>e2c6 8 £>g3 0-0 90-0 f5!? 10
c4 £>f6 11 cxd5 cxd5 12 £>c3 £>e4 13 £>ge2 a6
with an approximately equal position, but more
importantly, a position with some life.
6 £>e2 #f6 7 #xf6 £>xf6 8 £>bc3 £>bd7 9
0-0 0-010 lei Ie811 Ag5 g612 £>g3 4>g713
Ixe8 £>xe8 14 lei £>c7
V2-V2 Malaniuk-Khuzman, Lvov 1988.
o
4 c4 (D)

Exchange Variation
105
With this move, White signals that he may
not want a quiet game after all.
4...±b4+
I prefer this over 4...£tf6 but also there Black
is doing well; 5 £>c3 and then:
a) 5...c6 6 £tf3 ±d6 7 c5 £e7 (L.Acl 8
We2+ ±e6 9 £>g5 We7, as in Snape-R.Hall,
British corr. Ch 1993-5, is probably stronger) 8
JLd3 b6 (White maintains a small plus after the
less accurate 8...0-0 9 ±f4 b6 10 b4 a5 11 a3) 9
cxb6 axb6 10 0-0 ±a6 11 ±xa6 Sxa6 12 Wd3
0-0 13 ±g5 h6 14 ±h4 £>h5 15 ±xe7 Wxe7 16
Sfel ± M.Gurevich-P.Nikolic, Belgrade 1991.
b) 5..Ael 6 £tf3 0-0 7 cxd5 £>xd5 8 £d3
c5 9 0-0 £>c6 10 dxc5 £>xc3 11 bxc3 £xc5 12
Sbl h6 13 Wc2 Se8 = Plaskett-Lalic, British
League (4NCL) 2003/4.
5 £>c3 (D)
5 ±d2 ±xd2+ 6 £>xd2 £>f6 leads only to
equality.
iM±M*BSH
5...£>e7
This has been a popular way of meeting the 4
c4 Exchange Variation. The knight is somewhat
safer on e7 compared to f6 and not as
vulnerable to a JLg5 pin; it is also useful that it can later
attack d4 by moving to f5.
Nor is 5...!fe7+!? a bad idea. Then Miezis-
Tyomkin, North Bay 1999 continued 6 Ae2
±g4 7 cxd5 £>f6 8 ±g5 £>bd7 9 h3 Ah5 10 g4
JLg6 with a messy position that shouldn't be
unpromising for Black.
6£>f3
A key strategic point in Black's set-up is the
possibility of gaining control of important light
squares, d5 and c4 in particular. This was
clearly emphasized in Speck-Luther,
Liechtenstein 1993 after 6 a3?! £xc3+ 7 bxc3 0-0 8 £>f3
£>bc6 9 ±d3 dxc4 10 ±xc4 £>d5 11 ±xd5 (11
±d2 2e8+ 12 ±e2 We7 13 c4 £>b6 14 ±e3
^.g4 15 0-0 Sad8 also looked fairly comfortable
for Black in Santo-Roman - Shaked, Cannes
1997) 11...2e8+ 12 £e3 Wxd5 13 h3 Wc4 14
Wc2b6+.
6...±g4 7±e2
Watson gives 7 a3 £xc3+ 8 bxc3 £>bc6 9
£g5 0-010 JLe2, assessing 10...f6 11 ±d2 £>a5
as pleasant for Black, who will soon take over
control of the light squares, or 10...Wd7 11 0-0
dxc4 12 JLxc4 £}d5, again intending ...£>a5 and
...b5.
7...dxc4 8 ±xc4 0-0 9 ±e3 £>bc6 10 0-0
£>f5 11 Wd3 (D)
This position arises very logically. I find it
hard to believe that White should have any
advantage whatsoever but Miezis happily aims
for it almost any time he has a chance.
Il...£>d6
Black's main problem is that he has three
minor pieces for only two squares, viz. two
bishops and a knight to occupy d6 and f5. I guess
this is what Dvoretsky has called 'the
superfluous piece'. The knight and dark-squared bishop
form the most serious problem as both would
stand well on d6.
ll...Wd7 12 £>g5 £>a5 13 ±b5 c6 14 ±a4
±h5 (14...f6 15 £>ge4 £>xe3 16 fxe3 ±e6 17
Sf4 ±) 15 £>e2! ±g6 (15...Ae7 16 £>f4 ±xg5
17 £>xh5 £>xe3 18 fxe3 Sae8 19 Sael ±) 16

106
FRENCH: ADVANCE AND OTHER LINES
£rf4 b5 17 £>xg6 hxg6 18 ±c2 was better for
White in Miezis-Rodgaard, Gausdal 2003.
12 ±d5 ±f5
Stronger than 12...&e7 13 ±g5 h6 14 ±xe7
#xe7 15 Sfel Wf6 16 £>e5 Af5 17 Wf3 c6 18
±b3 Sad8 19 Se3 2fe8 20 Sael with the better
position for White, Miezis-S.Pedersen, Gausdal
2000.
13 #dl (D)
13...&e7
Other moves favour White:
a) 13...£>a5 14 ±g5 #c8 15 #a4 ±xc3 16
bxc3 £>c6 17 £>e5 £>e4 18 £>xc6 bxc6 19 ±xe4
±xe4 20 f3 ±d3 21 Sfel ±/± Miezis-Glek,
Hamburg 1995.
b) 13...±xc3 14 bxc3 £>e4 15 ±xc6 bxc6
16Wcl #d5 17c4#a5 18£>e5#a6 19f3£>d6
20 d5 (20 c5!?) 20...cxd5 21 cxd5 Sfe8 22 Af4
± Miezis-M.Gurevich, Bonn 1995.
14 ±b3 (D)
Pill A A A ±
mm. W
m itiii
«» ^1 Br IP M m%
14...c6
A fairly natural move but far from Black's
only option. Both 14...£ie4!? and 14..Jtg4!?
15 h3 ±h5 16 ±g5 c6 17 Sel Se8 with
approximate equality come into consideration.
15 £>e5 4>h8 16 Sel
16 Sel f6 17 £>d3 ±xc3 18 bxc3 £>d5 19
£>f4 £>xe3 20 fxe3 tfe8 was pleasant for Black
in Waitzkin-Shaked, Bermuda 1997 due to the
backward e-pawn and a strong potential
blockade on e4.
16 £ie2 is an idea, following up with ^g3.
16...f6 17 £>d3 ±xc3 18 bxc3 £>d5 19 c4
£>xe3 20 Sxe3 Se8
Black is doing fine, Miezis-Moreno Ruiz,
Andorra 2001.
D)
4 £>f 3 (D)
This is White's most flexible option but also
has the drawback that Black can pin the knight
with ...Jtg4 at some point. We shall later
discuss a few move-order issues associated with
the move.
Now:
Dl: 4...±g4 106
D2: 4...£>c6 108
D3: 4...£>f6 109
D4: 4...±d6 111
Dl)
4...±g4

Exchange Variation
107
This and 4...£}c6 tend to lead to the most
exciting positions. Obviously, White can react
passively and then nothing big is going on, but
there is also a sharp approach...
5 h3 ±h5 (D)
W
mm 'a mm mm"£ mm '"£
mm * mm HH * Wti *
ABAB BA
6#e2+
A sort of critical test of Black's 4th move,
introduced by Kasparov in 1991. Naturally, 6
Jte2 is harmless.
6...#e7
A point of White's last move - and #e2+
ideas generally in similar lines - is that on
6...±e7?, 7 #b5+ wins a pawn: 7...&c6 (or
7...£>d7 8 #xd5 £>gf6 9 #b3 ±) 8 #xb7 (8
GteSlT) 8..AM 9 #b5+ *f8 10 #e2 ±g6 11
£ia3±d6 12£ie5±.
7 ±e3 £ic6 8 £ic3 0-0-0
8...±xf3 9 #xf3 £ixd4 10 Wdl c5 11 £ixd5
tfe5 12c3! 0-0-0 13 cxd4cxd4 14#xd4#xd4
15 ^.xd4 Hxd5 16 JLc3 is a line given by
Kasparov that is certainly better for White due
to the two bishops.
9 g4! ±g6 10 0-0-0 (D)
10...f6
Good and solid. A sharper line is 10...^b4,
when after 11 a3 £>xc2 12 Af4,12...£ial 13 b4!
a5 14 #xe7 ±xe7 15 £>e5! worked out well for
White in Ulybin-Vilela, Santa Clara 1991, since
the al-knight is trapped (15...axb4 16 axb4
JLxb4 17 <i>b2 +-). However, Watson points
out that 12...^xa3! is far from clear. Black
gains a lot of pawns after 13 bxa3 #xa3+ 14
#b2 ±b4 15 #xa3 ±xa3+ 16 *d2 £tf6 and
shouldn't be worse. His main variation (I have
omitted a few deviations) runs 17 ^e5! JLb4!
18 f3 c5 19 £ixg6 hxg6 20 *c2 c4 21 £g2 b6
22 Ibl! £xc3 23 *xc3 *b7 24 Ihel *c6 25
Ie2 Ide8 26 Ieb2 Ie7 =. Therefore White
might want to consider the idea 11 tfoel £tf6 12
a3 £ic6 13^g2 £ie4 14 £ixe4 ±xe4 15 f3 £g6
16Wf2.
ni.g2
11 a3!? #d7 12 £id2 f5 13 £ib3 £tf6 14 f3
£d6 15 #d2 with an edge for White,
Kasparov-Short, Tilburg 1991.
U...ttd7 12 a3 (D)
■ilfl il
mm
12...£ige7
12...£\a5!?-Watson.
13 ±f4 *b8 14 £ia4 £ic8
14...b6 = Shirov.
15 £ih4 kSl 16 £>f5
Now instead of 16...b6 17 £ic3 £>8e7 18
£>e3 ± Mitkov-Shirov, Batumi Echt 1999, Psa-
khis suggests 16...£i6e7!? 17 £ic5 #c6 «>.

108
French: Advance and Other Lines
D2)
4...£>c6 (D)
As against 4 JLd3, this is a way to side-step
symmetrical piece-play, and avoids committing
Black to ...Ag4 yet.
5±b5
This is the consistent reply. We now have a
mirrored Winawer line (3 ^hc3 k.b4 4 exd5
exd5) with the extra tempo £}f3 for White.
Other moves:
a) 5 h3 M5 6 £d3 ±xd3 (Dreev suggests
6...Wd7 intending ...0-0-0 as a sharp
alternative) 7 Wxd3 ±d6 8 £>c3 £>ge7 9 £>b5 a6 10
£>xd6+ Wxd6 11 0-0 0-0 12 k.62 Sfe8 13 lael
£}g6 = Ivanchuk-Dreev, Wijk aan Zee 1996.
b) 5 ±d3 and now:
bl) 5...£g4 6£e3£d67&M2&ge7 8 0-0
h6 9 a3 0-0 10 c4 dxc4 11 £xc4 £>g6 = Con-
quest-Glek, Bundesliga 1996/7.
b2) 5...±d6 6 0-0 £>ge7 7 c4 dxc4 8 £xc4
0-0 (D) and now White should decide if it is
worth preventing... JLg4:
b21) 9 h3 ±e6!? (a noteworthy solution, but
a 'normal' move like 9...^.f5 would be slightly
better for White after 10 £>c3 Wd7 11 lei) 10
±xe6 fxe6 11 £>c3 (11 Wb3 Wd7 12 Wxb7?
Iab8 13 Wa6 Ixf3! 14 gxf3 £>xd4 with a strong
attack for Black - Polgar) ll...We8 (Polgar
reckons 1 l...Wd7 <*> is stronger) 12 lei Wg6 13
Se4! Iae8 14 We2 £>d5 15 Sg4! Wf5 16 ±h6
If7 17 £>h4 £>xc3 18 bxc3 Wa5, J.Polgar-
Kramnik, Dos Hermanas 1997, and now Polgar
likes White's attack after 19 Icl or 19 Wd3.
b22) 9 £>c3!? (queried by Polgar because
Black can pin the knight) 9...±g4 10 h3 ±h5
(10...£xf3 11 Wxf3 £>xd4 12 Wxb7 ±) 11 g4!
±g6 12 a3 *h8 <*> Pires-Carson, e-mail 1999.
b3) 5...&M!?6Ab5+!?(6Ae2Af5 7&a3
£k6! = V.Zaitsev-Volkov, Moscow 1999, and
others) 6...Ad7 7 We2+ ±e7 8 ±xd7+ Wxd7 9
£>e5 We6 (9...Wc8 10 a3 £>a6 11 £>c3 ± Milov-
Barsov, Germany tt 1995/6) 10 Wb5+ c6 11
Wxb7 Id8 12 0-0 £>xc2 13 £>xc6 Id6 14
Wb8+ ±d8 15 £>xd8 Ixd8 16 Wb5+ Id7 17
Wb8+Id8=.
5...±d6 6 c4
6 0-0 £>e7 7 c4 dxc4 8 £>c3 0-0 9 £xc4 - 5
Ad3 kd6 6 0-0 £hge7 7 c4 dxc4 8 kxc4 0-0 9
Zhc3.
6...dxc4 7 d5 a6 8 ±a4
8 JLxc4 gives Black no problems; for
example, 8...We7+ 9 ±e3 £>e5 10 £>xe5 Wxe5 11
£>c3 £>f6 ?.
8...b5 9 dxc6 bxa410 0-0 £>e7 11 Wxa4 0-0
12 £>bd2 (D)
The latest attempt for White in this line,
although Black seems to have at least two lines
that lead to equality. Instead, 12 Wxc4 ±e6 13
Wc2 Af5 14 Wa4 ±d3! gave Black enough
compensation in Kasparov-Bareev, Paris rpd
1991 to persuade most players to look for
something else.
12...Ib8
This might be safer than 12...c3 13 bxc3
We8, and now:
a) 14 £>e4 Wxc6 (14...Ab7? 15 £>f6+! +-)
15 Wxc6 £>xc6 16 £>xd6 cxd6 V2-V2 Kosash-
vili-Dolmatov, Haifa 1995.

Exchange variation
109
b) 14 £>d4 £b7 15 £>c4 £>xc6 16 £>xd6
cxd6 17 £>xc6 Wxc6 = Luther-Vallejo Pons,
Havana 2001.
c) 14 Aa3!? Wxc6 (14...±xa3 15 Wxa3
£>xc6 16 Sfel Wd8 17 Wc5 Wf6 =) 15 Wxc6
£>xc6 16 ±xd6 cxd6 17 £>c4 Sd8 18 Sfdl d5
19 £>b6 Sb8 20 £>xd5 £e6 21 £>e3 2xdl+ 22
£}xdl and White is a pawn up although it is
difficult to exploit, Dannevig-Brynell, Oslo
2004.
13 a3 Sb5!?
13...c3 14 bxc3 Sb6 15 Sel! £>xc6 16 £>c4
Sb5 17 ^xd6 cxd6 18 c4 was better for White
in Chandler-Bareev, Hastings 1991/2.
14 £>xc4 £>xc6 15 ±g5 f6 16 ±h4 We8 17
2felWf718£>xd6cxd6
= Fressinet-Bauer, Enghien-les-Bains 2001.
D3)
4...£>f6 (D)
The similarity to the Petroff is clearly seen
here, especially given that the position can arise
from the timid 1 e4e5 2 Zhf3 Zhf6 3 Zhxe5 d6 4
thf3thxe4 5d3thf6 6d4d5.
5±d3
On 5 Ag5 Black should play 5...&e7 with a
likely transposition to 5 k.d3 kel 6 0-0 0-0 7
$Lg5. In Ulybin-Bareev, Russian Ch (Elista)
1995, Black played 5...h6?!, but then Bareev
points out that White can gain a large advantage
by 6 We2+ ±e6 7 Axf6 since 7...Wxf6 8 Wb5+
wins a pawn.
5...±d6
The symmetrical strategy is a safe
continuation for Black but at some point it is clear that
Black will be forced to deviate, as a white threat
will come first. This is the only argument that
might entice White into playing these
positions, but in reality I think Black equalizes
fairly easily with sensible play. There are three
noteworthy alternatives:
a) 5...c5 had a brief period of popularity in
the early 1990s as an attempt to force more
exciting positions, but is now considered to give
White good chances of some advantage. 6 0-0
c4 7 2el+ £e7 8 £f 1 0-0 (D) with two options
for White:
al) 9 ^.g5 and now:
all) 9...Ag4 10 h3 Axf3 (10...Ah5 11 g4
±g6 12 £>e5 ± Kasparov) 11 Wxf3 £>c6 12 c3
£>e8!? 13 ±xe7 £>xe7 14 £>a3 £>c7 15 £>c2 ±
Kengis-M.Gurevich, Tilburg 1992.
al2) 9...h6 10 ±h4 Af5 11 £>c3!? (11 b3
cxb3 12 axb3 £>c6 13 c3 a6 14 £>bd2 Se8 15

110
FRENCH: ADVANCE AND OTHER LINES
£>e5 ±) ll...£>c6 12 £>e5 Se8 13 b3!? g5 14
±g3 Wa5 15 Se3 £>xd4 (15...£>xe5 might be
better, but White maintains an advantage with
16 Sxe5) 16 £>xc4! dxc4 17 #xd4 £>g4 18
Sxe7! Sxe7 19 bxc4! Sd8 20 £>d5 +- Rublev-
sky-Bashkov, Cheliabinsk 1993.
al3) 9...£>c6 10£>e5 ±e6 11 £>xc6bxc6 12
b3 c5 13 bxc4 dxc4 14 £>d2 h6 15 ±f4 cxd4 16
Jtxc4 = Hamdouchi-Komarov, Cannes 1992.
a2) 9 b3 cxb3 10 axb3 ±g4 (10...£>c6 11
£>bd2 ±f5 12 c3 a6 13 £>e5 ±) 11 h3 ±h5 12
g4! ±g6 13 £>e5 #c7?! (13...£>c6 14 £>xc6
bxc6 15 ±f4 Ab4!? 16 c3 ±xbl 17 cxb4 ±g6
18 f3! ± Bliumberg) 14 £>a3! a6 15 ±f4 with a
substantial advantage for White, Bliumberg-
Bruch, Schwabisch Gmiind 1994.
b) 5... Jte7. Now, in these, and similar,
positions White can elect to prevent a ... Jtg4 sortie
by playing h3 here or on the next move, but
Black will just do the same, and White hasn't
gained much. 6 0-0 0-0 7 ±g5 ±g4 (7...h6 8
Jth4 £fo5, as Barsov has played, is probably
also just equal) 8 £>bd2 £>bd7 9 c3 c6 10 Sel
Se8 11 #c2 ±h5 12 £>e5 ±g6 13 ±xg6 hxg6
14 #b3 Wcl 15 £>xd7 #xd7 = Krivokapic-
Malakhatko, Sozina 2004.
c) 5... Jtg4 (D) and now:
cl) 6 h3 ±h5 7 #e2+ #e7 (on the natural
7...±e7, White's idea is 8 g4 ±g6 9 ±xg6 hxg6
10 #b5+ £>bd7 11 #xb7, and now 1 l...Sb8 12
#xa7 £>xg4 13 £>c3 ±b4 14 Af4 #f6 15 #xc7
turned out clearly in White's favour in Lupu-
Schwartzman, Val Thorens 1992) 8 ±e3 #e6 9
£>bd2 £>c6 10 c3 ±d6 11 0-0-0 0-0 12 g4 ±g6
13 £>h4 ±xd3 14 #xd3 Sfe8 = Makarychev-
Nikolenko, Russian Cht (Moscow) 1994.
c2) 6 0-0 ±e7 7 Sel 0-0 8 ±g5 £>bd7 9
£>bd2 Se8 10 c3 h6 11 ±h4 £>h5 12 ±xe7 Sxe7
13 #b3 «tf4 14 Sxe7 #xe7 15 Sel WdS 16
±bl £>b6 = Roder-Khachian, Abu Dhabi 2000.
6 0-0 0-0 (D)
7h3
With the obvious purpose of preventing
... Jtg4. Other moves:
a) 7 Sel ±g4 8 Ag5 - 7 &g5 k,g4 8 UeL
b) 7 £>c3 Se8 8 ±g5 c6 9 Sel Sxel+ 10
#xel £>bd7 11 £>e2 £>f8 12 £>g3 £>e6 13 ±h4
g6 = Larsen-R.Byrne, Havana OL 1966.
c) 7 Jtg5 Jtg4 and now:
c 1) 8 £>bd2 £>bd7 9 c3 c6 (unless either side
tries something more ambitious, this should of
course just be equal; 9...c5?! 10 #c2 h6?! was
such an attempt, in D.Gurevich-Ivanchuk, Biel
IZ 1993, when White got a slight plus with play
against the isolated d-pawn after 11 dxc5! £>xc5
12 Ah4) 10 #c2 #c7 11 h3 ±h5 12 Sfel Sfe8
13 Sxe8+ Sxe8 14 Sel Sxel+ 15 £>xel ±g6
16 ±xg6 hxg6 17 £>ef3 Af4 18 Axf4 #xf4 19
g3 Wf5 20 #xf5 gxf5 = Spangenberg-Zarnicki,
Buenos Aires 1999.
c2) 8 Sel £>bd7 9 £>bd2 c6 10 c3 #c7 11
#c2 ±h5 12 ±h4 ±g6 13 ±xg6 hxg6 14 ±g3
Jtxg3 15 hxg3 Sae8 = Liogky-Baklan, France
2002.
7...Se8
Having got to the e-file first, Black will now
be the one to exchange rooks if White plays Sel.
After 7...h6 8 Sel Se8, White initiates the

Exchange Variation
111
exchange, and even though that is no big deal,
Black then has to take some care to equalize: 9
Sxe8+ Wxe8 10 £>c3 a6 (10...c6 cannot be so
bad either) 11 ±e3 £>c6 12 a3 £>e7 (12...±e6 is
probably also OK, but the text-move is
definitely better than 12...±d7?! 13 £>h4! #e6?!
14 ±f5 #e8 15 Wf3 with an advantage for
White, Chandler-Short, London 1986) 13 £>h4
£>g6! 14 £>xg6 fxg6 15 Wf3 Wf7 16 Af4 Axf4
17 #xf4 g5 18 #e5 ±e6 = PH.Nielsen-
L.B.Hansen, Tastrup 1992.
8 £>c3 c6 9 ±g5 h6 10 ±h4 £>bd7
Not much is going on, and the position is just
equal. A few more moves from Marinkovic-
Lucke, Bundesliga 1992/3: 11 #d2 #c7 12
Sael Af4 13 Sxe8+ £>xe8 14 Vel £>df6 15
£>e2 ±d6 =.
D4)
4...±d6 (D)
This may also transpose elsewhere but here
we shall focus on set-ups where Black plays his
king's knight to e7.
5±d3
A seemingly more ambitious approach is 5
c4, and then:
a) 5...!re7+ 6 ±e2 and Black can win a
pawn on c4, but White will get it back and in a
few moves Black's queen will prove poorly
placed on e7:
al) 6...±b4+7£>c3£>f6 8 0-0dxc49±xc4
0-0 10 Sel #d6 11 £>e5 with a slight
advantage for White, Illescas-Yusupov, Barcelona
1989.
a2) 6...dxc4 7 0-0 £rf6 8 ±xc4 0-0 9 Sel
#d8 10 £>e5 ± Marinkovic-Ivanovic,
Podgorica 1993.
b) 5..Af6 6 ^c3 0-0!? (6...c6-4 c4£>/6 5
Zhc3 c6 6 &J3 ±d6) 1 cxd5 h6!? (avoiding the
pin by Jtg5, and consequently keeping the
pressure against d5, anticipating to win back
the pawn) 8 ±e2 (8 £>e5 £>bd7 9 £>xd7 #xd7
10 ±e2 b6 11 0-0 ±b7 12 ±b5 Wf5 and White
found nothing better than repeating the position
with 13 ±d3 Wd7 14 ±b5 V2-V2 in Groszpeter-
Lputian, Budapest 1996) 8...£fod7 9 0-0 £>b6
10 £>e5 £>bxd5 11 #b3 c6 12 ±c4 ±c7 13 Sel
±b6 14 ±e3 ±e6 = Skripchenko-Zhukova,
Krasnoturinsk worn 2004.
5...£>e7 (D)
This is a good and sensible set-up and there
is really little reason to go into detail with it.
Black intends to develop his bishop with ... Jtf5
and simply equalizes.
6 0-0 0-0 7 ±g5
There are also other options for White at
this stage but against most Black will just play
...Jtf5, equalizing immediately.
7...f6 8 ±h4 £>bc6 9 c3 JkJtS 10 ±g3 W&7
= Arencibia-Mellado Trivino, La Pobla de
Lillet 1997.

8 King's Indian Attack
1 e4e6
When playing the French you will face a
King's Indian Attack set-up quite regularly at
any level. For many players it is just very useful
to have their opening scheme laid out in a
concrete pattern that is offered by a set-up like the
King's Indian Attack.
In order to avoid this chapter becoming
disproportionately large, against the King's Indian
Attack I shall offer only a repertoire-based
coverage from Black's point of view.
My choice of main line has fallen upon a
fianchetto development which maybe isn't very
French-like but nonetheless offers dynamic
play and shouldn't be more difficult to learn
than any other main line against the King's
Indian Attack.
We shall split our material into:
A: 2Ve2 112
B: 2d3 117
A)
2^e2(D)
This move is often associated with Chigorin.
One purpose behind it is that White may play 3
exd5 in reply to 2...d5, and then Black would be
forced to recapture with the queen. Thus Black
often replies differently from the way he would
meet 2 d3.
2...c5
Including 2...d5, Black has a number of other
options, which I'll briefly line up:
a) 2...b6 3 £>f3 Ab7 4 g3 c5 5 Ag2 £>c6 6
0-0 £>ge7 7 b3 d6 8 i.b2 e5 9 c3 g6 10 d4 i.g7
11 Sdl ®b8 12 £>a3 0-0 13 Sabl a5 14 £>c4
Aa6 15dxc5bxc5 16lrc2Sd8 17 Acl Wc7 18
£>e3 ± Lobron-Hertneck, German Ch (Bremen)
1998.
b) 2...5if6 3 e5 (of course, White may also
refuse to be provoked and play, for example, 3
d3 or 3 £>f3 instead) 3...£>d5 4 £>c3 (4 c4 £>f4 5
We4 £>g6 6 £>f 3 d6 7 exd6 &xd6 8 d4 c5 9 £>c3
cxd4 10 &xd4 fodl was at least OK for Black
in Motwani-Nogueiras, Istanbul OL 2000)
4...£>xc3 5 dxc3 d6 6 Af4 dxe5 7Wxe5!7 c6 8
Ad3 £>d7 9 We2 ?ic5 10 0-0-0 £>xd3+ 11
Sxd3 ®a5 12 <&bl looked promising for White
in Zimina-Bistrikova, St Petersburg worn 2001.
c) 2...e5 3f4!?(3&f3&c6-2...&c63£y3
^5)3...exf4 4^f3fDj.
Now, in a King's Gambit, White has the
extra move ®e2, but it is far from clear that this is
in fact an advantage, since it blocks the natural

King's Indian Attack
113
path for White's light-squared bishop and is
badly placed in front of the king:
cl) 4...g5 5 Ito!? ±e7 6 ±c4 c6 7 ®b3 d5
8 exd5 g4 9 £>e5 b5 10 ±e2 ±c5 11 d4! Wh4+
12 <A>dl f3 13 dxc5 fxe2+ 14 <A>xe2 £>f6 with an
unclear position, Brustkern-Cooke, Budapest
2002.
c2) 4...d5!? 5 exd5+ ±e7 6 d4 £>f6 7 ®b5+
(7 £ic3 0-0 8 ±xf4 Ab4 was fine for Black in
A.Sokolov-Hjorth, Copenhagen jr Wch 1982)
7...£>bd7 (7...c6 8 dxc6 £>xc6 9 Axf4 0-0 looks
like good compensation for Black) 8 Jtxf4 0-0
9 Jtd3 and White is better, Gallagher-Gerber,
Cannes 1999.
d) 2...£ic6 (D) and then:
dl) 3 £tf3 e5 4 c3 £>f6 (4...d5 5 d4!?) 5 d3
a5 6 Wc2 ±c5 7 ±e2 d6 8 0-0 0-0 9 £>bd2 h6
with an equal position, Bruzon-Nogueiras,
Havana 2003.
d2) 3 f4!? d5 4 exd5 £>d4 5 Wd3 (if 5 Wdl,
Black can recapture with the pawn) 5...®xd5 6
£>c3 Wd7 (6...Wd8!?) 7 £>f3 £>xf3+ 8 ®xf3
£>e7 9 b3 £>f5 10 ±b2 ±e7 11 ±d3 £>d4 12
Wg3 ±f6 13 0-0-0 ± McDonald-Pein,
Wrexham 1995.
d3) 3£>c3e5 4f4!?.
e) 2...d5 3 exd5 Wxd5 4 £>c3 Wd8 5 g3 c5
(5...£>c6!?; 5...±d7 6 ±g2 ±c6 7 £>f3 £>d7 8
b3! ±) 6 ±g2 £ic6 (Oil claimed Black should be
equal after 6...£>e7) 7 ±xc6+! bxc6 8 £rf3 £ie7
9 £>e4 £>f5 10 c3 f6 11 g4! £>e7 12 d3 e5 13
Sgl £id5 14 g5 f5? (14...±e6 is preferable,
although 15 gxf6 gxf6 16 Jte3! is good for White)
15£>xe5!fxe4 16dxe4£>c7(16...£>e7 17»c4!)
17 Wh5+ g6 18 £>xg6 hxg6 19 Wxg6+! <A>d7 20
±f4 We8 21 0-0-0+ £>d5 22 Wxe8+ <A>xe8 23
exd5 cxd5 24 g6! ± Oll-Glek, Netherlands
1997.
f) 2...ite7 has often been quoted as a way
of getting to play ...d5 and being able to
recapture with the pawn, but is of course
inappropriate if Black wants to play a fianchetto line
against a normal white set-up. 3 b3 (after 3
£rf3 d5 4 d3 £>f6 5 g3 we are out of our
repertoire) 3...d5 4 ±b2 Af6 5 e5 ±e7 6 ®g4 Af8.
This is the old game Chigorin-Tarrasch, St
Petersburg (14) 1893, which has been mentioned
again and again in the books. Despite the
massive loss of tempi with the bishop, Black's
position is rock-solid and not worse; e.g., 7 £tf3
c5 8 ±b5+ ±d7 9 ±xd7+ ^xd7 10 £>c3 £>c6
11 0-0 £>ge7 =.
We now return to 2...c5 (D):
White has a choice between two similar
setups:
Al: 3f4 114
A2: 3g3 116
The former features an f-pawn advance,
while the latter aims at more rapid
development. In both lines White has the option of
playing £\c3 at some point, with a close
resemblance to the Closed Sicilian. This is not very
common, though, since White's queen has then
been placed rather prematurely and not ideally
on e2.
Of course, 3 *hf3 £>c6 4 g3 is just a
transposition to 3 g3 Qsc6 4 Qsf3.

114 French: Advance
Al)
3 f4 £>c6 4 £>f3 (DJ
4...d6
In my opinion, this is Black's best set-up. He
intends a fianchetto of the dark-squared bishop
and natural development with ...£\ge7, ...0-0 and
maybe advancing on the queenside. Compared
with lines arising from a Sicilian, White's
queen has been committed rather early to e2.
Other options include:
a) 4...£>ge7 and here:
al) 5b3!?d5 6^b2b6 7d3d4 8£>bd2g6 9
g3 k,gl 10 Ag2 e5 11 0-0-0 #c7 ¥ Fedorov-
Feygin, Minsk 1997.
a2) 5 g3 d5 6 d3 g6 7 Ag2 Ag7 8 0-0 b6 9 e5
£>f5 10 Wf2 h5 = Bellini-Zaja, Reggio Emilia
2000/1.
b) 4...i.e7 5 g3 d5 6 d3 £>f6 7 Ag2 0-0 8
0-0 b5, Kovacs-Uhlmann, Sarajevo 1969, is
another good line for Black. In general Black
tends to be quicker on the queenside in such
positions.
5g3
The game Lotzien-Keitlinghaus, Bundesliga
2002/3 took a rather unusual path: 5 d3 g6 6
Ad2 k,gl 7 Ac3 e5! 8 £>bd2 £>ge7 9 fxe5 dxe5
10 #f2 £>d4 and Black was already doing well.
5...g6 6 c3 k,gl 1 k,gl £>ge7
This is often arrived at via other move-orders;
e.g., 3 g3 foc6 4 c3 Zhge7 5 ±g2 g6 6 fit ±g7 7
£hf3d6.
8 0-0 0-0 (D)
9&a3
and Other Lines
This prevents an immediate ...b5 advance
from Black and keeps the option of later going
d4 in one move. The main alternative is 9 d3 but
White cannot reckon with any advantage after
this:
a) 9...e5 10 Ae3 exf4 11 A.xf4 h6 12 £>bd2
(12 £>a3!?) 12...d5 13 <4>hl ±e6 = Rigo-Keit-
linghaus, Dortmund 1991.
b) 9...b5 10 Jte3?! (according to Dolmatov,
this is already wrong; White should first play
10 a3 a5 and only then 11 Jte3, in order to avoid
Black's next move) 10...b4 11 2c 1 (11 #c2
#a5 12 £tfd2 Aa6 13 £>b3 #b5 14 c4 #b6 15
£Md2 Jtb7 16 £>f3 a5 = Guseinov-Nikolenko,
USSR Cht (Azov) 1991) ll...bxc3 12 bxc3
Aa6 13 £>bd2 Sb8 14 i.f 1 Wd7, Morozevich-
Dolmatov, Krasnodar 1997. It is clear that
Black does not have any problems; his pieces
are more harmoniously placed and he is ready
to attack White's centre with ...e5 and ...f5.
c) 9...b6!? 10 £>bd2 a5 11 Sdl a4 12 a3
±a6 13 £>h4 Sc8 14 £>df3 £>a5 15 Ae3 £>b3
16 Sabl d5 17 e5 d4 18 cxd4 cxd4 19 A.f2 Wd7
T Erenska-M.Socko, Lubniewice 2002.
9...Sb8 (D)
10*hl
Since White prepares to play d4 later, this
may be a useful move, as the king is off the a7-
gl diagonal. Other moves:
a) 10 e5 d5 (10...f6!?) 11 d4 cxd4 12 cxd4
£>f5 13 £>c2 h5 (another idea could be 13...#b6
14 b3 f6) 14 Ah3 Ah6 15 b3 <4>h7 16 Aa3 Sh8
17 Sacl was better for White in Belkhodja-
Vareille, Paris 1994. With the rather inflexible
pawn-structure that arises after 13...h5, Black

King's Indian Attack
115
has little counterplay to compensate for White's
space advantage.
b) 10 £>c2 b5 11 d3 b4 12 ±62 a5 (Black
could also play without this, because the a5-
square is then vacated later on for, say, the
queen; 12...jta6 is a good alternative) 13 labl
#b6 14 £>e3 Aa6 (14...bxc3 15 £>c4 #b4 16
Axc3 Axc3 17 bxc3 #xbl 18 Ixbl Ixbl+ 19
*f2 «) 15 c4 £>d4 16 £>xd4 ±x64 17 <4>hl is
slightly better for White, Kiik-Biriukov, Imatra
1999.
We now return to 10 *hl (D):
10...a6
Or:
a) 10...d5 Ile5#a5(ll...b6!?) 12£>c2b5
13 a3 d4?! 14 cxd4 cxd4 15 £>fxd4 £>xd4 16
£>xd4 ±bl 17 b4 Axg2+ 18 #xg2 #b6 19
Jtb2, Fedorov-Zakharevich, Minsk 1997. Black
has some compensation for the pawn but not
quite enough.
b) 10...Wd7 11 d4 (this should be prepared
by 11 Idl, according to Gleizerov) ll...cxd4
12 cxd4 £>xd4 13 £>xd4 ±x64 14 £>b5 ±gl\
(14...i.c5?! 15 b4!? or simply 15 Idl is good
for White) 15 Idl d5 16 £>xa7 #a4! 17 £>b5?!
(17 £>xc8 Ifxc8 18 e5 is better, with
approximate equality) 17.. Jtd7 18 £>c3 Jtxc3! 19 bxc3
Ab5 20 #el dxe4 21 Id4 #a6 ? Fedorov-
Gleizerov, Vilnius 1997.
c) 10...f5!? 11 d3 b5 12 exf5 £>xf5 13 ±62
d5 14 g4 £>h6 15 £>g5 W61 16 lael b4 17 £>bl
bxc3 18 bxc3 £>d8 19 Jte3 « Lastin-Bareev,
Russian Ch (Elista) 1997.
Il£>c2
11 Idl b5 (11. ..e5 and ll...f5 are both worth
considering) 12 d4 b4?! 13 dxc5! bxc3 14 £>c4
d5 15 £>b6 Wcl 16 bxc3 dxe4 17 #xe4 Axc3
18 Ibl was better for White in Arakhamia-
Grant - Hagarova, Leon worn Echt 2001.
11..A5 12 d4 (D)
12...cxd4
12...b4?! (again, this turns out badly) 13 dxc5
bxc3 (13...dxc5!? 14 Idl Wcl is probably
better) 14 b3!? dxc5 15 Ae3 £>d4? (Psakhis
gives 15...Wa5! 16 #c4 £>b4 17 Ifcl ±bl 18
Axc5 £>xc2 19 Ixc2 Ifc8 20 b4 #a4, with
counterplay, as Black's best) 16 £>fxd4 cxd4 17
ladl e5 18 fxe5 £>c6 (18...A.xe5!? 19 ±\64
Wcl 20 #e3! and White rounds up the c-pawn)
19 £>xd4 £>xe5 20 £>c6 (but not 20 £>e6? ±xe6
21 Ixd8 Ifxd8, and Black has a lot of
compensation) 20...£>xc6 21 Ixd8 Ixd8, Fedorov-Yan-
demirov, St Petersburg 1994. Simplest is now
22#c4!±.

116
FRENCH: ADVANCE AND OTHER LINES
13 cxd4 b4!
13...d5 14 e5 is slightly better for White.
14 b3 a5 15 kb2V.
White's bishop would be better placed on e3,
although 15 Jte3 f5!? (for example) isn't bad
for Black.
15...1rb6 16 Sfcl Aa6 17 Wdl d5! 18 e5
Sfc8
+ Onishchuk-Shaked, Tilburg 1997. Black
has excellent play on the queenside, with ...a4
and ...£\a5 coming, or ...£\a7-b5.
A2)
3 g3 £>c6 4 £>f3 (D)
4 c3 is seen frequently but only as a
transposition to other lines; e.g., 4...g6 5 kg2 Jtg7 6 f4
(or 6 £>f3 £ige7 7 0-0-4 *hf3 g6 5 kg2 kg7 6
0-0 *hge7 7 c3) 6...£>ge7 7 £>f3 d6 - 3f4 *hc6 4
G%3 d6 5g3g6 6 c3 kg7 7 kg2 Zhge7.
Obviously White may play an early c3 in a variety of
these lines.
4...g6
In this line we shall concentrate on Black
fianchettoing his bishop and mostly avoiding
...d5.
5 kg2 kg! 6 0-0
Some attempts at playing quickly on either
the queenside or the kingside, and thus
delaying castling, have been tried but do not impress;
e.g.:
a) 6 d3 £\ge7 7 c3 e5! ? (it can be quite useful
for Black to have the option of going ...d5 in one
move, and this grabs space; in the following the
d-pawn is even left untouched for surprisingly
long) 8 a3 0-0 9 b4 cxb4 10 axb4 b5 11 0-0 h6
12 Ae3 f5 13 £>a3 f4 14 gxf4 exf4 15 Ad4 g5!
16 kxgl (Glek gives 16 Wa2+ <4>h7 17 kxgl
<A)xg7 18 £ixb5 g4 followed by ...f3 with a
mess) 16...*xg7 17 c4! g4 18 1^2+ <4>h7 19
cxb5 gxf3 20 Axf3 £>g6! 21 bxc6 dxc6 with
Black having perhaps more than adequate
compensation, in view of his promising attack, Oni-
shchuk-Glek, Bundesliga 1996/7.
b) 6 c3 £>ge7 7 h4 h6 8 d3 e5 9 a3 a5 10 a4
d6 11 £ia3 Jte6 = Hall-Schmittdiel, Bundesliga
1998/9.
6...£>ge7 7 d3
Should White, by 7 c3 0-0 8 £ia3, try to carry
out d4 in one move, then Glek thinks Black is
even better after 8...e5! 9 Sdl d5!, when White
is slightly misplaced compared to a regular
King's Indian.
7...0-0
7...d5 -2d3d5 3 Me2 c5 4 ?hj3 *hc6 5 g3
Zhge7 6k.g2g6 7 0-0k.g7.
8 c3 (D)
8 £>c3 is a standard Closed Sicilian set-up
except that White has played his queen to e2,
while on 8 £>bd2 Black can develop with ...d6,
...e5/...b6/...b5, etc.
8...e5!?
We have already come across this move in
some of the notes up to here. The lost tempo
with the e-pawn is of course a disadvantage for
Black but the d4 break is prevented and Black
might want to carry out ...d5 in one move, when
the character of the position takes on that of a

King's Indian Attack
117
normal King's Indian fianchetto line. We shall
consider three alternatives:
a) 8...d5 9 e5 (D) and now:
al) 9...b5 10 h4!? h5 11 ±g5 Wc7 12 £>bd2
fcf5 13 Sfel a5 14 £>fl Sb8 15 £>lh2 with
promising chances for White, Psakhis-G.Mohr,
Istanbul OL 2000. White is threatening g4 with
a serious attack.
a2) 9...f6!? 10 exf6 Sxf6 11 ±e3 b6 12
£>bd2 a5 13 Sfel h6 14 Af4 g5 15 ±e5 £>xe5
16 £>xe5 Sf5 17 £>df3 1U6 18 Sadl h5 19 h3
Jtf6 is unclear, Bologan-de la Riva, France
2002.
b) 8...b6!?9h4h610£>bd2±b7 1l2eld6
12 £>h2 b5 13 £>df3 &h7 14 ±e3 b4 and
Black's play is going well on the queenside,
while White's attack on the kingside isn't
convincing, Smyslov-Kurajica, Bar 1980.
c) 8...d6 and now:
cl) 9£>a3 2b8!10£>c2b5 11±f4?!(lld4
b4 12 ±d2 bxc3 13 ±xc3 cxd4 14 £>cxd4
£>xd4 15 £>xd4 Wb6 =) ll...b4 12 d4 bxc3 13
bxc3 cxd4 14 £icxd4 (14 cxd4 e5 ?) 14...£ia5 ?
Lagvilava-Malakhov, Minsk 1997.
c2) 9^ibd2Sb8(9...h610Sbllrc7 1lSdl
b6 12 £>f 1 e5 13 £>e3 ±e6 14 b4 cxb4! 15 cxb4
b5 ¥ Smyslov-Cramling, Roquebrune (Ladies
vs Veterans) 1998) 10 £>b3!? b5 11 d4 cxd4 12
£>fxd4 b4 13 ±g5 Wb6 = Tosic-Sedlak, Budva
2003.
9a3d6
9...a5 10 a4 d5 11 exd5 £>xd5 12 £>fd2 h6 13
£>a3 ±e6 14 £>dc4 Sb8 15 £>b5 Wd7 = Eren-
burg-Grafl, Budapest 2004.
10 b4 h6 11 ±b2 ±e6 12 b5 £>a5 13 £>bd2
2c8
= Kaminski-M.Socko, Koszalin 1999.
B)
2 d3 d5 (D)
White now needs to prevent a capture on e4
followed by a queen exchange. We consider
two ways for him to do so:
Bl: 3^2 117
B2: 3£>d2 123
Bl)
3!fe2
This is not quite as popular as 3 £>d2 but
remains a very flexible line, where White's
knight may take different routes. For the sake of
giving the reader more choice, I give two main
replies for Black:
Bll: 3...dxe4 117
B12: 3...c5 120
Bll)
3...dxe4 4 dxe4 (D)
4...e5
First and foremost this is a very solid system
for Black, who easily develops and keeps the
structure symmetrical. White pins his hopes on
a very slight lead in development and maybe
slightly the better manoeuvring prospects. If
we count tempi, White also has the extra move

118
FRENCH: ADVANCE AND OTHER LINES
#e2 but this is often not very beneficial since
he will have to move the queen again in order to
develop his bishop.
The main alternative is 4...b6, when White
should take into account that Black wants to
play ...Jta6 next. There are a number of
options:
a) 5 £>d2! ? ±a6 6 £>c4 £rf6 7 £tf3 £>c6 8 c3
±e7 (8...£>a5 9 #c2 ±xc4 10 ±xc4 £>xc4 11
#a4+ £>d7 12 #xc4 £>c5 13 <4>e2! ±) 9 e5 £>d7
10 #e4 ±b7 11 #g4 g6 12 ±h6 b5! 13 £>e3
£>dxe5 14 £>xe5 £>xe5 15 ±xb5+ (15 #h3!?)
15...C6 16 #e4 Wcl 17 ±e2 f5! 18 #a4 £>f7 oo
Morozevich-Lputian, Wijk aan Zee 2000.
b) 5 £tf3 ±a6 (DJ and now:
W
■a
p
a A n wfi A a
bl) 6 c4 ±b4+!? 7 £>bd2 £>c6 8 a3 ±e7 9
b4 £>d4 10 £>xd4 #xd4 11 Sbl ±g5 12 £>b3
#c3+ 13 ±d2 ±xd2+ 14 #xd2 #xd2+ 15
<4)xd2 e5 = Strikovic-Drasko, Vrnjacka Banja
1999.
b2) 6 #d2 #xd2+ 7 ±xd2 £>f6 8 £>c3
±xfl 9 Sxfl £>c6 10 0-0-0, Morozevich-Vol-
kov, Samara 1998, and now Morozevich gives
10...0-0-0! 11 £>g5 Sd7 12 ±e3 £>d8 (the
alternative 12...Jtd6!? also comes into
consideration) 13 f3 h6 14 £>h3 £>c6 =.
c) 5 c3 ±a6 6 #c2 ±xf 1 7 <4>xf 1 (D). Now:
...J ■Ail A
§p Hf a if"
®*k
/MmmAmim
7// 0777?/,
cl) 7...£>d7 8 £tf3 £>gf6 (this seems more
accurate than the immediate 8...£>c5) 9 Jtf4?!
(probably already a mistake; 9 e5 £>d5 10 £>bd2
#c8 should be about equal) 9...£>c5 10 e5
Wd3+ 11 #e2 £>d5 12 ±g3 0-0-0 13 b4 #xe2+
14 <4)xe2 £>e4 15 2c 1 g5! and White is in
trouble, Fedorov-Ghane Gardeh, Dubai 2002.
c2) 7...±d6 8 £>f3 £>d7 9 ±e3 £>gf6 10 h3
#c8 11 £>bd2 #a6+ 12 c4 0-0 13 <4>e2 e5 14
Shel Sfe8 15 *f 1 #b7 = Murdzia-Bunzmann,
Griesheim 2004.
We now return to the position after 4...e5
(D):
W
AAA mm * H *
A
AS A 'BttB AH
8f^jll§

King's Indian Attack
119
5 £>f 3 £>c6 6 c3 «M6 7 #c2
The last few moves are logical, with both
sides developing naturally, and White now
improves the position of his queen, preparing to
develop the light-squared bishop. It wouldn't
be worth considering a fianchetto, since Black
could then take advantage of the bishop's
absence from the f l-a6 diagonal by playing
something like ...a5 followed by ...b6, with ...Jta6
coming.
7...Ad6
There are a great many alternatives:
a) 7...±c5?! 8 b4 ±b6 9 ±e2 0-0 10 0-0
±g4 11 £>bd2 a5 12 b5 £>e7 13 £>c4 ± Sve-
tushkin-Khrushchov, Minsk 2000.
b) 7...a5 8 ±b5 ±dl 9 0-0 ±c5 10 ±g5 #e7
11 £>bd2 0-0 12 Sadl h6 13 ±h4 g5 14 ±g3
£>h5 <*> Beshukov-Doroshkevich, Cherkessk
1997.
c) 7...a6 (D) and now:
cl) 8 ±e2 ±d6 9 0-0 0-0 10 £>bd2 h6 11
£>c4 b5 12 £>xd6 cxd6 13 Sdl #c7 14 a4 Sb8
15 axb5 axb5 16 h3 Jte6 was roughly equal in
Naumann-Vokler, German Ch (Heringsdorf)
2000.
c2) 8 b4 ±d6 9 £>bd2 0-0 10 £>c4 (the same
plan as in our main line is applied but here
Black will obtain counter-chances against the
white queenside) 10...h6 11 Jte2 (11 a4!? might
be more accurate to pre-empt the following)
ll...b5!? 12 £>xd6 cxd6 13 a4 ±b7 14 0-0 (14
axb5 axb5 15 Sxa8 #xa8 16 ±xb5 Sc8! is
certainly not bad for Black) 14...*c7 15 ±d3
£>e7! 16 lei Sfc8 17 ±b2 d5 18 exd5 £>exd5,
Svetushkin-Kruppa, Kiev 2000, and now White
should probably play 19 axb5!? axb5 20 Sxa8
Jtxa8 21 Jtxb5, although Black has enough
counterplay for the pawn with 21...#b8!?.
d) 7...£g4 8£b5!?£xi3 9£xc6+bxc610
gxf3 ±c5 11 ±g5 (11 £>d2 £>d7 12 £>c4 Wf6
13 #e2, intending Jte3, appears more sensible
to me) 1 l...h6 12 Jth4 (D) and here:
dl) 12...g5!? 13 ±g3 £>d7 14 £>d2 Wf6 15
0-0-0 0-0-0 16 £>c4 Ab6 17 #e2 #e6 <*> Mai-
wald-Beckhuis, Berlin 1998.
d2) 12...*d6 13 £>d2 £>h5 14 0-0-0 #e6 15
#b3 £rf4 16 ±g3 Axf2 17 #xe6+ fxe6 18
Axf4 (18 *c2 ±xg3 19 hxg3 £>g6 20 £>b3 ±)
18...exf4 19 £>c4 Sd8 20 2xd8+ 4>xd8 21 £>e5
was better for White in A.Sokolov-Bauer,
French Cht 1994.
We now return to 7... Jtd6 (D):
8£>bd2

120
FRENCH: ADVANCE AND OTHER LINES
The immediate direction of the knight to c4
seems best, though White has also tried other
things:
a) 8 ±g5 h6 9 ±h4 ©e7 (this keeps options
open since Black may consider castling queen-
side; 9...g5 10 ±g3 £>h5 oo) 10 £>bd2 ±e6 11
Ab5 (11 £>c4 g5 12 ±g3 0-0-0 =) 1 L..Ad7 12
^fl a6 13 JLe2 JLc5 was approximately equal
in Beshukov-Poliakov, Krasnodar 2001.
b) 8 ±e2 0-0 9 0-0 £>h5!? 10 £>bd2 £>f4 11
£>c4 £>xe2+ 12 «xe2 ±e6 13 Sdl «e7 14 b3
Sfd8 15 JLb2 f6 = Vescovi-RotSagov, Moscow
2002.
8...0-0 9 £\c4 (D)
9 ±e2 Se8 10 0-0 £>h5 11 g3 ±h3 12 Bel
Wf6 13 £>c4 £>f4 14 Axf4 exf4 15 Badl Af8
16 e5 Hi6 V2-V2 Fedorov-B.Socko, Batumi Ech
2002.
9...H6
Having castled, it is worth preventing JLg5.
Allowing this would give White a good game:
a) 9...We7 10 ±g5 h6 11 ±h4 ±e6 12 ±e2
a6 13 Bdl Bfd8 14 b4!? ± Rabiega-Sielecki,
German Ch (Bremen) 1998.
b) 9...Se8 10 ±g5 ±e6 11 Bdl «e7 12
±e2 Bad8 13 0-0 Bd7 14 b4 ±xc4 15 ±xc4 is
much better for White, Godena-Mariano, Bratto
1999.
10£e2Bb8
Preparing ...b5. Another option is 10...We7,
when White should probably castle and play
b4, with perhaps an edge.
11 a4 b6 12 0-0 ±b7 13 Bel £>a5 14 £>xd6
cxd6 15 ±fl «c7 16 b4 £>c6
16...£>c4 17 £>d2 £>xd2 18 ±xd2 also seems
slightly better for White.
17±a3
White is better, Zhang Zhong-Speelman,
Bled OL 2002.
B12)
3...C5 (D)
W
im a m
HI A
HP ^^ A HP
lAjwi m.
A
w& wm/k'-m^. m
n
m
ABABWHAH
4£>f3
Compared to 2 ©e2 d5, the line 4 exd5
©xd5 5 £\c3 ©d8 is now more favourable for
Black, and certainly not a problem.
4...£>c6 5 g3 £>ge7 6 ±g2
An example of White's flexibility in the
King's Indian Attack is 6 h4!?, which is
obviously directed against Black's intended fian-
chetto set-up. However, I doubt it is a line that
will ever become very popular, and Black
reacts well with 6...e5! (D).
W

King's Indian Attack
121
The lost tempo with the e-pawn doesn't
matter because White's move with the h-pawn may
turn out more harmful to himself, as the g4-
square is weakened. Black now simply adopts a
set-up similar to that of the Samisch Variation
in the King's Indian. Then:
a) 7 exd5 £>xd5 8 £>xe5? is bad in view of
8...£>d4!.
b) 7 £>bd2 f6 8 c3 ±g4 9 ±h3 ±xh3 10
Sxh3 «d7 11 Shi d4 12 c4 £>c8 13 <4>fl ±e7
14 h5 £>d6 15 <4>g2 0-0-0 = Certic-J.Horvath,
Yugoslav Cht (Niksic) 1997.
c) 7 ±g2 ±g4 (7...f6 8 £>a3 ±e6 9 c3 «d7
10 0-0 0-0-0 11 &e3 d4 12 cxd4 cxd4 13 Ml
4^8 t Zhang Zhong-Movsesian, Zagan jr Wch
1997) 8 c3 f6 9 £>bd2 «d7 and Black already
has a comfortable position. The game Gen-
occhio-Sax, Montecatini Terme 2000 went 10
£>f 1 d4 11 £>lh2 ±e6 12 cxd4 £>xd4 13 £>xd4
«xd4 14 0-0 £>c6 15 ±e3 «d7 ?.
6...g6 7 0-0 kgl (D)
8c3
White would like to close the centre with e5,
and this is a preparation for that as White can
then follow up with d4. The immediate 8 e5 is
feasible as well and might transpose to our main
line, but Black can also try to exploit the fact that
White has omitted c3. That is by 8...h6
(preparing ...g5) 9 h4 *c7 10 Sel (10 ±f4?! g5! 11
hxg5 hxg5 12 Axg5 £>xe5 t) 10...£id4!? 11
£>xd4 cxd4 12 c4 (D) (12 c3 £>c6 13 ±f4 g5 14
«g4 £>xe5 15 Axe5 ±xe5 16 cxd4 ±f6 was
good for Black in Glueck-Damljanovic,
Philadelphia 1987) and now:
a) 12...dxc3 13 £>xc3 Ad7 14 ±f4 «b6 15
©d2 ^if5 = Maiwald-M.Gurevich, Groningen
open 1997.
b) 12...dxc4 13 dxc4 (13 £>a3!? cxd3 14
£>b5 «a5 15 £>d6+ <4>f8 16 fd2 «xd2 17
±xd2 oo) 13...±d7 14 £>d2 ±c6 15 ±xc6+
£lxc6 16 £M3 d3! and Black was doing fine in
Psakhis-Khenkin, Khalkidhiki 1992.
8...b6 (D)
It is considered advisable for Black to wait a
little before castling, as White will then close
the centre and play for a kingside attack. By
aiming to develop the queenside first, Black
stays more flexible and often intends to castle
in the other direction.
8...0-0 - 2 Me2 c5 3 g3 *hc6 4*hf3g65 ±g2
±g7 6 0-0 Zhgel 7 d3 0-0 8 c3 d5.
9e5
White doesn't have a very useful way of
waiting with this any more. On 9 Sel h6 (9...d4!?)

122
French: Advance and Other Lines
10 h4 Black can opt between 10...Jk.a6 11 e5
Wc7 (transposing to 9 e5 h610 h4 Wc711 2e7
ka6) or 10...dxe4!? 11 dxe4 e5 12 £>a3 0-0 13
£>h2 f5 14 Sdl ±a6! 15 Wc2 We8 16 £>b5
±xb5 17 Wb3+ <4>h7 18 Wxb5 f4 • Zhang
Zhong-Vescovi, Linares 2001.
9...H6
The h-pawn moves are often thrown in at an
early stage. Black prepares ...g5, and White
should prevent this, which is positionally
beneficial for Black as he has better control of the
dark squares.
10 h4 ±b7
Or 10...ffc7 11 Sel ±a6 (D) with the idea of
preventing d4 and intending to castle queen-
side, but this appears slightly risky as White
can prepare b4-b5. Often Black seeks counter-
play by sacrificing a pawn on the kingside with
...g5 but in the majority of lines this does not
appear a fully satisfactory solution:
a) 12 a4!? g5?! (maybe 12...£>a5 13 £>bd2
£tec6 with the idea of ...g5 as in Salvador
Lopez-Delchev, Cullera 2001) 13 hxg5 hxg5
14 ±xg5 £>g6 15 £>a3 £>cxe5 16 £>b5 £xb5
(16...fto7 17 £>xe5 ±xe5 18 f4 Ab8 19 f5
£\f8 20 Wf3 ± Iordachescu-Filippov, Moscow
1996) 17 axb5 <4>f8 18 £>xe5 ±xe5 19 Wf3
<4>g7 20 Sa4 ± Cabrilo-Ye Jiangchuan, Novi
Sad OL 1990.
b) 12 £>a3 0-0-0!? 13 Sbl (13 Ad2,
preparing b4, is also good, but it is worth noting that
13 JLf4?! here and in many similar lines is the
wrong square for the bishop because of 13...g5
14 hxg5 £lg6!, with strong counterplay). Now:
bl) 13...*b8 14b4cxb4 15cxb4b5 16^c2
Sc8 17 a4 £>xe5 18 £>fd4 g5 19 hxg5 hxg5 20
±xg5 £>7g6? (20...bxa4 21 b5 ±b7 doesn't
look clear) 21 axb5 ±xb5 22 £>xb5 Wxc2 23
tfe3 ± Bologan-Khrushchov, Minsk 2000.
b2) 13...g5 14 hxg5 £>g6 (14...hxg5 15
±xg5 £>g6 16 b4 £>cxe5 17 £xd8 Wxd8 18
£>xe5 £>xe5 19 c4 ±) 15 gxh6 ±xe5 16 ±g5
Sdg8 17 Wd2 ±d6 <*> Lastin-Rotsagov,
Moscow 1996.
ll2elWc7 12£>a3a6(T>J
13£>c2
13 Af4 £>f5 14 Sacl We7 15 d4 cxd4 16
cxd4 0-0 (16...£>cxd4 17 £>xd4 £>xd4 18 We3
WM 19 £>c4! dxc4 20 ±xb7 Sd8 •) 17 £>c2
Sac8 18 Wd2 <4>h7 = Nijboer-Lautier, Breda
1998.
13...2c8!?
Again this is directed against d4 since after
an exchange White cannot recapture with the
pawn. Other moves:
a) 13...g5?! 14 hxg5 £>g6 15 gxh6 ±xh6 16
±xh6 Sxh6 17 d4 0-0-0 18 dxc5 bxc5 19 We3
Sh5 20 Wxc5 £>cxe5 21 Wxc7+ <4>xc7 22 £>xe5
£lxe5 23 Sadl ± Lobzhanidze-Dgebuadze,
Schwabisch Gmund 2001.
b) 13...d4 14 cxd4! cxd4 15 £>cxd4 £>xd4
16 £>xd4 Sd8 (16...±xg2 17 <4>xg2 0-0 18 ffe4
is also much better for White) 17 JLxb7 Wxbl
18 £tf3 Sd7 19 We4! ± Adams-M.Gurevich,
Sarajevo 2000.
14 d4
After 14 Af4, 14....£tf5 15 Sacl! a5 16 d4
0-0 17 Wd2 We7 18 Afl was good for White

King's Indian Attack
123
in Bologan-Van Wely, Romanian Cht 2000.
Bologan offers 14...g5! as a significant
improvement, giving Black a good attack after
15 hxg5 £>g6 16 gxh6 £>xf4 17 gxf4 Axh6 18
£>g5 We7 19 Wh5 ±gl 20 Wg4 2g8.
14...cxd4 15 £>cxd4
This is slightly better for White according to
Bologan, although any advantage must be very
small after 15...^xd4 16 £>xd4 £>c6.
B2)
3 £>d2 c5 4 £>gf 3 £>c6 5 g3 g6 6 Ag2 ±g7 7
0-0 (D)
This position may come about from several
move-orders. The fianchetto lines lead to very
flexible positions for both sides, and the often
very amorphous nature of the positions makes
them difficult to cover theoretically, but I hope
the following pages stand out as a good
overview of ideas for both sides.
Black would usually prefer e7 for his knight,
although it is not clear that f6 is much worse.
Generally, Black tends to postpone castling, as
White can then close the position with e5 and
subsequently has a clear target on the king-
side. Instead, Black often plays ...b6 intending
... Jta6/Jtb7, waiting for White to show his
intentions, which he in turn also attempts to hide
for a while.
We shall look at:
B21: 7...d4!? 123
B22: 7...£>f6 125
B23: 7...£>ge7 126
While the third option is clearly the most
common, the first two deserve attention and are
good lines if one wants to avoid the most
theory.
B21)
7...d4!? (D)
W
HP ASP P
mM
Black often closes the centre in this way, but
usually he only does so after White has played
Eel, for example.
8a4
Natural and good. White prepares ^c4
without having to worry about ...b5 - a common
King's Indian strategy. A more direct line was
seen in Svetushkin-Bursteinas, Oropesa del
Mar U-18 Wch 1998: 8 £>b3 ffe7 (8...b6 9 e5
Jtb7 10 Sel £>ge7 led to White's advantage in
Vasiukov-Holm, Polanica Zdroj 1972, although
he had to reposition his b3-knight) 9 e5!?
(White's last move would have been out of
place if Black managed to play ...e5 himself)
9...£>xe5 10 £>xe5 Axe5 11 £>a5 ±c7 12 c3
itf8 (White has reasonable compensation after
12...Axa5 13 Wa4+ Wd7 14 Wxa5 b6 15 Wa3
Ab7 16 ±xb7 Wxb7 17 cxd4 cxd4 18 Af4) 13
Wa.4 (White has already spent lots of time with
his knight, but 13 £>b3 would nonetheless make
some sense as Black's pieces are oddly placed)
13...Sb8 14 cxd4 Ad7 15 tfa3 cxd4 16 b4 <*>.
8...e5 9 £>c4 (D)
9...W.
According to Sakaev, 9...^ge7 is less
accurate due to 10 c3 0-0 11 cxd4 exd4 (1 l...cxd4 12

124
French: Advance and Other Lines
WW/ A Wfc WB A W& A
— ^H
b4 would win White space on the queenside or
a vital central pawn) 12 JLf4 JLe6 13 £>g5 ±.
With the text-move, Black strengthens the
centre and prepares harmonious piece deployment
with ...£\h6-f7 and ...JLe6.
10 c3
10 £>el £>h6 11 f4 £>f7 12 £>f3 (T>) seems
like a normal King's Indian plan but Sakaev
criticizes White's 10th move.
Now:
a) 12...0-0 13 f5 leads to a slight advantage
for White.
b) 12...JLe6 looks natural, but Sakaev
rejects it in view of 13 f5: 13...JLxc4 14 fxg6
hxg6 15 dxc4 ±, 13...±d7 14 fxg6 hxg6 15
£>h4, or 13...gxf5 14 £>h4 fxe4 15 ±xe4 with
compensation.
c) 12... JLf8! (this move is difficult to
understand at first, but maybe it is clever
prophylaxis) 13 Sf2 ±e6 14 ftfl (the main point of
Black's 12th move is that 14 f5 is now met by
14...±d7 and if 15 fxg6 hxg6 16 £>h4, Black
has simply 16...Sg8) 14...«d7 15 fxe5
(according to Sakaev, other moves lead to Black's
advantage: 15 f5 gxf5 16 £>h4 f4! 17 gxf4 £>d6 ?;
15 £>h4 exf4 16 gxf4 ±e7 ?) 15...fxe5 16 £>g5
£>xg5 17 JLxg5 JLg7 and now, rather than 18
g4? £>d8! 19 Af6 Axf6 20 Sxf6 ±xc4 21 dxc4
£te6 with clearly the better chances for Black
due to his much superior knight against the
bishop (Va.Chernov-Sakaev, Chisinau 1998),
White should play 18 Sf6! Axf6 19 Wxf6 Sf8
20 £>xe5 Wd6 (20...We7?! 21 Wxe7+ £>xe7 22
JLh6 Sg8 23 £tf3 with good compensation for
White) 21 £>c4 Wd7 22 £>e5 with equality
(Sakaev).
10...£ih6 11 cxd4 cxd4 12 a5 £>f7 (D)
w
, wmJL
mm mm
w§ mm wm a m
13 b4
Direct; White could also have chosen to
prepare it with 13 JLd2.
13...£>xb4 14 Wa4+ £>c6 15 a6 0-0 16 ±a3
He8
Now:
a) 17 Sabl (it is quite natural to put the
rooks on the b- and c-files, but it turns out to be
ineffective) 17...Wc7 18 Sfcl (18 axb7 ±xb7
19 Wb3 £>a5! 20 £>xa5 Wxa5 21 Sfcl ±c8 +)
gives White some pressure but in Demkovich-
Vysochin, Barlinek 2002 it turned out
insufficient after 18...Sb8.
b) White might be able to exert more
pressure with 17 axb7 ±xb7 18 Sfbl! Wc7 19 Wb3,
the difference being that White's rook on al
(rather than on fl) prevents 19...£ia5.

King's Indian attack
125
B22)
7...£>f6 (D)
W
For most of us, it would feel more natural to
put the knight on e7, but there is no doubt that
the knight is more influential on f6. One
drawback is that if White takes on d5 immediately,
Black is now more or less committed to
recapturing with the knight. Another is that White
may close the centre and try to build up an
attack on the kingside.
8c3
No hidden intentions here: White wants to
close the centre with e5 followed by d4. There
are other options:
a) 8 tfe2 0-0 (Lputian also suggests 8...b6!?
since closing the centre would then be dubious:
9 e5 £>g4 10 c4 0-0 +) 9 £>b3 b6 10 e5 £>d7 11
Af4 a5 12 c4 a4 13 £>bd2 a3 14 bxa3 Se8!
(more accurate than 14...Sxa3, when White
favourably repositions his knight with 15 £M
Sa5 16£ic3) 15Sacl ±b7 16£>bl d4 17h4h6
18 Sfel Sa7 19 Sc2 Wa8 and Black might even
be very slightly better, Svidler-Lputian, Erevan
1996.
b) 8 exd5 £>xd5 (8...exd5 9 Sel+ £>e7 10
tfe2 is awkward for Black) 9 £>b3 (9 £>e4 b6
and now White has several options like 10 c4,
10 Ag5, 10 a3, 10 h4 or 10 £>fd2, but none
impresses that much) 9...b6 (D) and now:
bl) 10 c4 £\de7 (10...£\c7 11 Af4! with the
idea of ll...±xb2 12 d4! ±xal 13 Wxal ±
Velickovic) 11 d4 ±a6 12 dxc5 ±xc4 13 Sel
Wxdl 14 Sxdl 0-0 15 £>fd4 £>xd4 16 £>xd4
W
Sad8 17 ±g5 Sxd4 18 ±xe7 Se8 19 ±d6 bxc5
(19...Sd8!?) 20 ±xc5 Sxdl+ 21 Sxdl ±xa2
22 JLxa7 JLb3 V2-V2 Ionescu Brandis-Zilber-
man, Bucharest 1997.
b2) 10 d4!? ±a6 11 Sel c4 12 £>bd2 0-0
(12...c3 13 £\e4 cxb2 14 JLxb2 <*> has to be
investigated; if Black can hold a blockade on c4
he will be doing well - if not, White gains lots
of space) 13 c3 Wc7 14 Afl £>a5 15 £>e5
Sad8 16 Wg4 ±c8 17 Wh4 b5 18 £>e4 and
White is better, Hamdouchi-Benitah, French
Cht2001.
c) 8 Sel 0-0 9 c3 (D) gives Black a choice
of continuations:
■JLtt ■#
sf'W ' '""
1 m
cl) 9...a5 10£sfl!?(10e5£id7 11d4cxd4
12 cxd4 f6 13 exf6 Wxf6 was good for Black in
Y.Geller-Poluliakhov, Krasnodar 2001) 10...a4
(10...dxe4 11 dxe4 Wxdl 12 Sxdl £lxe4 13
JLe3 and Black can hardly avoid having to
return the pawn) 11 e5 £>d7 12 M4 a3 13 b3 f6

126
FRENCH: ADVANCE AND OTHER LINES
14 exf6 £>xf6 (14...Wxf6!?) 15 fto2 £>h5 16
±h6 Wf6 17 ±xgl Wxgl * Bologan-Koma-
rov, Ulcinj 1997.
c2) 9...b6 10 e5 £>d7 11 d4 f6 12 ±h3!? (12
exf6 Wxf6 13 £>b3, Stein-Sokolsky, USSR
1960, might also give White an edge) 12...fxe5
(the passive 12...Se8 13 exf6 Wxf6 14 £>b3 is
now an even better version for White, but
12...f5!?oois feasible) 13 .&xe6+<4>h8 14±xd5
±b7 15 dxe5 £>dxe5 16 ±e4!? (16 Sxe5 ±xe5
17 ±xc6 ±xc6 18 £>xe5 ±b7 19 Wb3 Wf6 20
f4 Wf5 does give Black some counterplay but
could be better scrutinized) 16...Wf6 17 £>xe5!
Wxf2+ 18 <4>hl ±xe5 19 kgl Wf5 20 £>c4
Sad8 21 ffe2 kgl 22 ±f4, Kremenietsky-
S.Ivanov, USSR 1983. White has returned the
pawn but is better due to his stronger piece
deployment.
c3) 9...<4>h8!? 10 exd5 (10 We2 b6 11 e5
£>d7 12 d4 cxd4 13 cxd4 £>b4 <*> Lputian; 10 e5
£>d7 11 d4f6 12 exf6 Wxf6 is good for Black,
P.Reh-Glek, Bad Zwesten 2002, or Black could
exchange on d4 before playing ...f6) 10...exd5
(10...£>xd5 11 £>b3 b6 12 d4 c4 13 £>bd2 tfc7
<*> Belotti-Lputian, Reggio Emilia 1997/8, but
this type of structure is often to White's
advantage as he has reasonable chances of preparing
a kingside attack) 11 £>b3 b6 12 Af4 £>h5 13
±g5 f6 14 ±e3 f5 15 ±g5 fto6 = Khait-
Nikolenko, Moscow Ch 1987.
8...0-0 9 e5 £>d7 10 d4 (D)
There are basically two plans for Black in
this position. He can hope for the best on the
kingside and seize space on the queenside, or
he can go for counterplay against White's
centre. A blend of both is rarely possible.
10...b5
This obviously initiates a space-gaining
action on the queenside. Another interesting idea
is 10...f6!? 11 exf6 Axf6 (ll...Wxf6 appears
more natural, but Glek probably wanted to
avoid something like 12 £>b3 c4 13 ±g5 tff7
14 £lcl, when he cannot immediately follow
up with an ...e5 break) 12 dxc5 (after 12 £>b3
c4 13 £lbd2 e5 Black seizes the initiative)
12...£>xc5 13 £>b3 £>xb3 14 Wxb3 £>a5 15
Wdl e5 <*> lordachescu-Glek, Porto San Giorgio
2001.
llSelb412£>fl.ia6(T>)
It is interesting that Khalifman sees no reason
to move the a-pawn forward before developing
the bishop. Thus he will keep the a5-square free
for a piece. In Adams-Khalifman, Dortmund
2000, White kept a symbolic advantage but
nothing of any real substance: 13 h4 bxc3 14
bxc3 Wa5 15 ±d2 £>b6 (15...Iab8 16 c4 Wb6
is another idea) 16 c4 fk4 17 cxd5 Wxdl 18
Saxdl £>xd5 19 dxc5 Iac8 20 a3 Sc7 21 ^e3
£>ce7 22 £>xd5 £>xd5 23 ±b4 Ib8 24 Id4 Af8
25 Scl Sbc8 26 £>g5 ±xc5 27 Sxc5 Ixc5 28
±xc5 Ixc5 29 Ia4 £>c7 30 f4 Scl+ 31 <4>h2
Sc4 32 Sa5 Sc2 33 <4h3 h5 34 £>e4 ±e2! =
with enough counterplay to hold the position.
B23)
7...£>ge7 (D)
8 Bel

King's Indian Attack
127
The most interesting alternative to this is 8
exd5, after which 8...£ixd5 transposes to 7...^hf6
8 exd5 <5W5, and 8...exd5 9 d4 gives us:
a) 9...cxd4 10 £>b3 Wb6 11 ±f4 (11 ±g5
0-0 12 £>fxd4 £tf5 = Dvoretsky) 11...0-0 12
±d6 d3! 13 c3 Id8 14 ±c5 fa6 15 lei £tf5
16 ±fl b6 17 ±xd3 Wb7 18 ±d4 £>fxd4 19
£>bxd4 Jtg4 = V2-V2 Seul-Levin, 2nd Bundes-
liga 1996/7.
b) 9...c4!? 10 c3 0-0 11 lei Af5 12 £tfl
±e4 13 ±f4 h6 14 h4 f6!? 15 £>3h2 ±xg2 16
*xg2 Wd7 17 Wg4 fxg4 18 £>xg4 h5 19
£>gh2 <4>f7 = Mohaupt-Podzielny, Goch 1996.
8...b6 rz>;
With this Black doesn't yet commit himself
to castling kingside. This has a specific
advantage in the fact that if White closes the centre
immediately Black can counter with a timely
...g5. Alternatively, 8...0-0 9 h4 h6 10 e5
usually develops in White's favour:
a) 10...1rc7 11 We2 b5 12 £>fl b4 13 £>lh2
*h8 14 Af4 £>g8 15 £>g4 We7 16 Wd2 h5 17
£>gh2 ±d7 18 £>g5 £>d4 19 £>hf3 £>f5 20 ±h3
± Iordachescu-Prasad, Erevan OL 1996.
b) 10...f5 11 exf6 Ixf6 12 £>h2 If7 13 £>g4
fd6 14 £>b3! *h7 15 c4 ±d7 16 ±e3 b6 17 d4!
cxd4 (17...dxc4 18 dxc5 fxdl 19 laxdl £>d5
20 £>d4 £>cb4 21 c6 ±c8 22 ±e4 ± Oratovsky-
Maiwald, Vejen jr Ech 1993) 18 £>xd4 dxc4,
Cyborowski-Kiseleva, Legnica 1995, and now
19h5g5 20lcl±
From the diagram White basically has two
ideas. He may close the centre with e5 or
capture on d5. The first idea is often prepared by
playing c3 first, so that he can guard it with d4
IHJ.H+B ■
V/At///, ////////. //^////. - //AX//A -
afterwards. Besides, White may flick in h4 at
some point, trying to soften up the black king-
side or induce some sort of weakening, i.e.
...h6, which on the other hand can also help
Black since he gains better control of the dark
squares. There are some transpositions;
therefore ideas where White plays h4 are dealt with
together, but White may find that an early h4
works better in collaboration with closing the
centre.
Thus, we have:
B231: 9c3 127
B232: 9exd5 129
B233: 9h4 131
It is considered premature for White to close
the centre immediately. One example: 9 e5 Wc7
10 We2 g5 (the thematic way of meeting White's
e5) 11 £>xg5 fxe5 12 fxe5 ±xe5 (and here
Black should always capture with the bishop,
since 12...^xe5 13 £>c4! complicates) 13 c3
±a6 14 k£\ kgl 15 £>df3 0-0 16 h4 h6 17
£>h3 e5 and Black is doing well, Amin-Glei-
zerov, Linares 2003.
There is also 9 a3, intending Sbl and b4 but
this is hardly the biggest test: 9...h6 10 Sbl a5 11
h4 a4 12 exd5 exd5 13 £>fl 0-0 14 Af4 Ia7 15
c3 4>h7 = Ljubojevic-Petrosian, Milan (6) 1975.
B231)
9 c3 (D)
As the immediate 9 e5 is premature, this is
an important option, preparing the e-pawn
advance since it may now be supported by d4.

128
French: Advance and Other Lines
Gaining valuable space on the queenside,
while still postponing the decision to castle,
and providing a more natural harmony in the
development of the bishop to a6. Other moves:
a) 9...h6 and now:
al) I0h4-9h4h610c3.
a2) 10 exd5 exd5 11 d4 cxd4 12 £>xd4 £>xd4
13 cxd4 0-0 14 £>c4 £>c6 15 ±f4 ±e6 16 £>e5
£>xe5 17 JLxe5 Sc8 = Kochetkov-Moskalenko,
Moscow 1995.
a3) 10 e5 g5 11 d4 £>g6 12 h3 0-0 13 £>f 1
f5 14 exf6 fxf6 15 ±e3 cxd4 16 cxd4 ±d7
and Black was doing fine in Filipovic-Psakhis,
BanjaLukal985.
b) 9... JLb7 10 e5 and now:
bl) 10..Mel is answered by 11 d4 when, on
ll...cxd4 12 cxd4 £>b4, White can play 13
«a4+.
b2) 10...h6 11 d4 g5 12 £>fl cxd4 13 cxd4
b5 14 g4! f b6 15 £>g3 ± Nun-Antoniewski,
Czech Cht 2002/3.
b3) 10...g5 11 £>xg5 £>xe5 12 £>df3 £>5g6
13 d4 h6 14 £>h3 ftt7 15 a4 £>c6 16 £>f4 £>xf4
17 JLxf4 ± Yurtaev-Gulko, Moscow OL 1994.
c) 9...0-0 10 e5 (naturally White also has
other options but this is principled) 10..Mel 11
#e2 g5 (the thematic way of indirectly
attacking e5) 12 £>xg5 f xe5 13 £>de4!? (D) (leading
to a critical position and certainly better than 13
f4? Wxe2 14 Sxe2 JLa6, which was already
more or less winning for Black in Hoehn-S.Ped-
ersen, Duisburg U-18 Wch 1992) and here:
cl) After 13...£ig6, 14 £>xh7?! <4>xh7 15 f4
#c7 16 £>g5+ <4>g8 17 f h5 Sd8 18 f h7+ <4>f8
19 h4 thgel wasn't convincing for White in
Kochetkov-Moskalenko, Alushta 1994, but 14
f4 f c7 15 £>f2, as in Van der Weide-Weeks,
Wijk aan Zee 1997, retains some pressure.
c2) 13...dxe4!? and now 14 f h5 h6 15 ±f4
ff6 16 ±xe4 e5 17 £>h7 Wd6 18 ±g5 Sd8 19
±xe7 f xe7 20 ±xc6 ±bl 21 ±xb7 f xb7 was
winning for Black in Berry-Gershon, Dresden
2003. Instead after 14 M4 ff6,15 ±xe4 e5 16
±xh7+(?) <4>h8 17 £>e4 ffe6 18 Wh5 has been
given as winning for White in several sources,
although 18...Wg4! forces a rather favourable
queen exchange for Black! In this line, 15
£lxh7! i>xh7 16 JLxe4+ is more complicated
and far from clear.
We now return to 9...a5!? (D):
W
IflAM+H tr
liiii
§A a
AH B BAfl
mm m
10 a4
10 ^fl d4!? (10...a4 is a natural and good
alternative) 11 e5 and now ll...dxc3 12 bxc3
£>d5 13 ±g5 tW7 14 c4 £>c7 15 £>e3 h5 16

King's Indian Attack
129
£>c2 ±a6 17 fd2 Sb8 18 Sadl was very good
for White in Filipovic-Barlov, Novi Sad 1985.
However, Black's reaction was a little too
experimental. There is nothing wrong with 11...0-0
orll...±b7.
10...Aa6
10...h6 11 h4 (transposing to 9 h4 h6 10 c3
a5 11 a4) is probably Black's best option. In
general, with the h-pawn moves in, Black
benefits from better control of the dark squares on
the kingside.
11 exd5 (D)
mXmwm m
Hill
™ •mat, •
'""''Wi
myywB,* mm* mm
W/M 4Ws h A
AiflB BAH
b Up
17 a5 £>xa5 (17...b5 18 a6 ±a8 19 £>a3 £>f6 20
Wf3 b4 co) 18 £>xa5 bxa5 19 #g4 Shg8 20 #c4
£>b6 21 ±xb7+ <4>xb7 22 #e4+ <4>b8 <*> Neved-
nichy-Manik, Budapest Z 2000.
b) 10 d4 cxd4 (10...0-0 11 c4 £>de7 12 dxc5
bxc5 13 £>e4 ± S.B.Hansen-Moskalenko,
Copenhagen 1995) 11 £ib3 and now Black
actually has to be careful:
bl) ll...Ab7?! 12 £>fxd4 £>xd4 13 £>xd4
threatens c4, and after 13...Sc8 14 Sxe6+! (D),
the position is suddenly exploding:
ll...£>xd5
ll...exd5 12 £>b3 0-0 13 d4 c4 14 £>bd2 b5
15 axb5 ±xb5 16 b3 cxb3 17 £>xb3 a4 18 £>c5
was better for White in Paragua-Delchev, Turin
2000.
12^ic4 0-013h4lrc714h5
White is better, Kaidanov-Zapata, New York
1993.
B232)
9 exd5 (D)
9...exd5
Due to White's slight lead in development,
he is able to exert pressure on Black's centre no
matter how Black recaptures. Black may end up
with hanging pawns after this but on the other
hand keeps good control of the centre and has
active pieces.
9...£>xd5 and then:
a) 10 h4 ±b7 11 £>c4 h6 12 a4 Wc7 13 h5
gxh5 14 £>h4 0-0-0 15 c3 ±f6 16 Wxh5 ±g5
bll) 14...fxe6? 15 £>xe6 Wd7 16 £>xg7+
and here:
bill) 16...Wxg7 17 ±xd5 ±xd5 18 Wxd5
Wd7 19 We5+ *f7 20 ±h6 Shg8 21 Sel Sc5
(21...Sc6 22 Se4 g5 23 JLxg5 also looks very
difficult for Black) 22 Wf4+ Wf5 23 We3 (23
Wa4 might be even stronger) 23...Wd7 24 b4
(24 Wf3+!7 Sf5 25 Sdl fixf3 26 Sxd7+ <4>e6
27 Sxa7 ±) 24...Sc6 25 b5 Se6 26 Wf4+ Sf6

130
French: Advance and Other Lines
27 #c4+ Ie6 28 Be4 Ic8 29 If4+ *g8 30 h4
Hce8 31 a4 ± Komliakov-Moskalenko, Noia-
brsk 1995.
bll2) 16...*f7 17 ±h6 Ihd8 18 ffe2 *g8
19 We5 (19 c4 £>b4 20 ±xb7 Wxbl 21 b3 <*>)
19...£fo4? (a stronger defence is put up by
19...Ixc2 20 £ih5 Well 21 Wxcl Ixc7 22
Jtxd5+ ±xd5) 20 £\h5 with a winning attack
for White in Howell-Soln, Bled 1995, because
of20...gxh5 21±h3.
bl2) 14...£te7! was later found as an
improvement, with 15 Jtxb7 fxe6 16 Jte3 (16
±xc8? ±xd4 -+) 16...Hc4 given, the line
continuing 17 £>xe6 #xdl+ 18 ixdl =. However, I
don't think Black's problems are entirely over
after 17c3!;e.g., 17...»d7 18±a6lc5 19#b3
and White has plenty of pressure.
b2) 11...0-0! 12 £>fxd4 £>xd4 (12...±b7 13
£\xc6 ±xc6 14 £>d4 ic8 15 £>xc6 Sxc6 might
also be playable for Black, despite White's
bishop-pair and three vs two on the queenside;
the knight on d5 is very strong and severely
restricts White's dark-squared bishop) 13 ^xd4
±a6 14 c3 V2-V2 Psakhis-Dokhoian, Sochi 1988.
Obviously, there is plenty of play left in the
position. Also, 14 ^c6 should be examined but
Black has enough counterplay; e.g., 14...#d6
15 ±xd5 exd5 16 £>e7+ *h8 17 #xd5 #xd5
18£>xd5lfe8 19£>e3Had8.
10 d4 (D)
10 £tfl 0-0 11 c3 certainly does not trouble
Black. After 1 l...h6 12 h4 ±e6 13 Af4 Ic8 14
#d2 <4)h7 Black has reasons to be quite content
with his position, Lauridsen-Vasquez, Chile
10...0-0
10...cxd4 11 £>b3 is just good for White,
with play against an isolated pawn. 10...c4?!
requires a look but is probably just bad as well, in
view of 11 ^e5!:
a) ll...£>xe5 12 dxe5 ±e6 13 £>f3 £>c6 14
±g5 »d7 15 »d2 0-0 16 Af6 h6 17 ^d4 ±
Kislinsky-Palecha, Ukrainian U-16 Ch (Kiev)
2000.
b) ll...£>xd4 12 £idxc4 £>e6 13 ±g5!! f6
(13...£>xg5 14 £>c6 or 13...±b7 14 £>xf7 wins
for White on the spot) 14 £>c6! £>xc6 15 ±xd5
£to4 (15...^e7 looks more resilient but after
16 ±xa8 £>xg5 17 #xd8+ *xd8 18 Iadl+
±d7 19 ie2 Black can hardly survive) 16 ±xe6
(16 ±xa8!? #xdl 17 Saxdl fxg5 18 c3 +-)
16...#xdl 17 laxdl ±xe6 18 Ixe6+ *f7 19
Ie4 Ihd8 20 £>e5+ *g8 21 Ixd8+ Ixd8 22
Ixb4 fxg5, Pavlenko-Trubits, corr. 1988, and
now simply 23 £M3 gives White an extra pawn.
lldxc5bxc512£ib3(T>)
12...c4!?
With this Black voluntarily compromises
his own structure but in any case he cannot
keep his pawns where they are. I prefer White
after 12...#b6 13 c3 (13 ±e3 d4 14 ±g5 {or
14 Jtf4} is also interesting):
a) 13...Id8 14 ±e3 d4 15 cxd4 c4 16 d5
Wcl 17 ±f4 #b6 18 £>fd2! cxb3 19 dxc6
£>xc6 20 #xb3 ± Nevednichy-Roca, Manila
OL 1992.
b) 13...c414^bd4(14±e3!?#b715^bd4
looks good too) 14...£>xd4 15 £>xd4 ±xd4 16
cxd4 Jte6 17 b3!, once again with a substantial

King's Indian Attack
131
advantage for White, Oratovsky-Kiriakov,
VejenjrEch 1993.
c) 13...±g4 14 ±e3 d4 (14...±xf3 15 ±xf3
d4 16 cxd4 cxd4 17 ±g5 ±) 15 cxd4 cxd4 16
£>bxd4 Iad8 (16...#xb2 17 £>xc6 £>xc6 18
#a4 ±d7 19 ladl £ie5 20 Ixd7 £ixd7 21
Wxdl #xa2 22 £>g5 Iad8 23 #xa7 #xa7 24
±xa7 ± Oratovsky-Soffer, Tel Aviv 1993) 17
£>xc6 #xc6 18 We2 ^d5 19 lac 1 »d7 20
±c5 Ife8 21 ttt2 ± Gaponenko-Yashchenko,
Hradec Kralove girls Ech 1992.
13 £>bd4 ±g4 14 ±e3! (D)
14 h3 £ixd4 15 hxg4 £>e6 16 Ibl Ic8 17 b3
#a5 is unclear, Reinderman-Bosch, Dutch Ch
(Enschede) 1993.
u...mi
The pawn-structure is generally to White's
advantage. Hence Black must avoid too many
exchanges. Something like 14...Ib8 15 b3 Wdl
16 »d2 Ifc8 17 ladl Ib7 18 £>xc6 Ixc6 19
Jtd4 would be to White's advantage.
15#d2lfe816£>xc6?!
This may be premature; 16 fiadl Had8 17 b3
is better, keeping the tension in the centre.
16...£>xc6 17 £>d4 £>e5!
This is the problem. Black is now preparing
...Jk,h3, and if then Jthl Black proceeds with
...£>g4.
18±h6
It is worth noting that 18 f4? c3! 19 #xc3
Hac8 is very good for Black.
18...±h8
Kaiszauri-Tukmakov, Vilnius 1978. The
position is difficult to assess but in view of his
more active pieces Black may be slightly
preferred.
B233)
9 h4 (D)
9...H6
White must not be allowed to play h5; now
this is met by ...g5. But in fact, it isn't clear who
fools whom with these h-pawn moves. With
White moving h4, Black is usually permitted to
reply ...h6. The other way around comes to the
same result: often Black starts with ...h6, which
White automatically meets with h4.
10 e5
White has a number of alternatives to this,
with 10 c3 being the most important, and of
course similar to 9 c3. Thus:
a) 10 *fe2 ±a6!? 11 e5 #c7 12 £>fl Ic8
(12...Ib8!?; 12...b5!?) 13 c3 b5 14 Af4 b4 15
£Mh2bxc3 16bxc3#a5 17lacl c4 18d4lb8
19 £>g4 *d7!? 20 #d2 Q$£5 <*> Dzhumaev-
Barua, Raipur 2002.
b) 10 exd5 exd5 11 d4?! cxd4 12 £>b3 (the
insertion of the h-pawn moves makes this a
poor plan) 12...±g4 13 Af4 0-0 14 »d2 ±xf3
15 ±xf3 *h7 16 Be2 »d7 17 lael Bfe8 18
±g2 a5 19 a3 Iad8 ? Shirov-G.Hernandez,
Merida 2000.
c) 10 c3 (D) with a major branch:
cl) 10...Jtb7 is one natural possibility.
c2) 10...±a6 11 exd5 £ixd5 (ll...exd5 12
d4! ±) 12 £>c4 (12 #a4 ±b7 13 d4 cxd4 14
£ixd4 Wdl 15 £l4f3 Id8 =) 12...0-0 13 £>fe5

132
FRENCH: ADVANCE AND OTHER LINES
£>xe5 14 £>xe5 #c7 15 c4 ±xe5 16 cxd5 exd5
17 Jtxh6 Hfe8 with a roughly balanced
position, Titov-Adamski, Warsaw 1992.
c3) 10...dxe4 11 dxe4 ±a6 12 #a4 ±d3 13
Ie3!? (13 £>c4 Ib8!, with the idea 14 £>ce5 b5
15 £>xc6 bxa4 16 £>xd8 a3!, looks fine for
Black) 13...b5 14 #a6 (14 fdl seems more
natural, intending &el) 14...Hb8 15 £\el c4 16
£>xd3 cxd3 17 ±fl (17 £tf3!?) 17...Ib6 18 #a3
£te5 with an unclear position, Varavin-Moska-
lenko, Alushta 1994.
c4) 10...a5 (D).
W
This move became significantly important
(not only in this exact position but also in
similar ones) after Kasparov convincingly beat Lju-
bojevic in 1983. There are several ideas behind
it. First of all, the advance of the a-pawn may
attempt to disrupt White's queenside structure;
and should Black wish to develop his light-
squared bishop on a6, the position appears
more harmonious. Moreover, and perhaps the
most important feature, the a-pawn clears
Black's 2nd rank, making the rook move ...Ha7
feasible. The rook is then away from the a8-hl
diagonal, which in many lines makes ...d4 a
good idea. Besides the rook has useful purposes
on its 2nd rank, sometimes from c7 or d7. On
the other hand, the black king must normally
stay in the centre.
c41) 11 We2 Ia7 12 e5 h5 gave Black
typical counterplay in Khachian-Kiriakov, Sochi
1997.
c42) 11 £rfl dxe4 (1 l...a4 and 1 l...d4!? are
other sensible options) 12 dxe4 #xdl 13 fixdl
a414±f4!?(14a3±a615±f4e5 16±e3ld8
17 ixd8+ ^xdS 18 £Md2 sbcl was
approximately equal in Dorenberg-Emms, Gent 2002)
14...a3 15 bxa3 ±xc3 16 labl Ixa3 17 ±c7
±a5 18 ±xb6 ±xb6 19 Ixb6 Ixa2 20 £>e3 0-0
21 Icl Ia5 22 £>c4 Ia4 23 £>ce5 £>xe5 24
thxeS and White regains his pawn with a
minimal advantage, Gretarsson-Khenkin, Stockholm
1991.
c43) 11 a4 (D) and now:
c431) 11.. .Jta6 12 exd5. Black may now
recapture in two ways:
c4311) 12...exd5 13 £>b3 0-0 14 d4 c4 15
£>bd2 b5!? 16 £rfl b4 17 ±f4 (17 £>e5!?)
17...f6 18 ±h3 ±c8 19 ±xc8 Ixc8 ? Visser-
Van Wely, Haarlem 1995.
c4312) 12...£>xd5 13 £>c4 #c7 and after 14
£>a3?! Id8! 15 £>b5 ±xb5 16 axb5 £>ce7 17
±fl 0-0 18lfe2£>f5 19±d2lfe8 20ladl e5!
Black could be very satisfied with the outcome

King's Indian Attack
133
of the opening in Kholmov-T.Pahtz, Varna
1987. More critical is 14 h5!? g5 15 £>xg5 hxg5
16 Axd5 0-0-0! 17 Axc6 #xc6 18 Axg5 Sd7
(18...f6? 19 «fg4 Axc4 20 dxc4 fxg5 21 Sxe6
Wdl 22 ffe4 is winning for White) 19 We2
Jtb7 20 f3 oo. A more solid continuation is also
available in 14 Jtd2, followed by H>3 and, for
example, Sadl.
c432) 11...2a7!?(D).Now:
W
c4321) 12 £>b3?! d4! 13 cxd4 cxd4 14 Ad2?
(14e5! ±a6! 15 Se4Sd7 16 Af4<*> Kasparov)
14...e5 15 £>cl Ae6 16 Se2 0-0 and Black
already has much the superior position with
...f5 coming, Ljubojevic-Kasparov, Niksic
1983.
c4322) 12 £rfl d4 13 cxd4 cxd4 14 e5! £>b4
15 Sa3 Aa6 16 Se4 Sd7 17 £>lh2 and now
Black does not seem any worse after 17.. JLb7,
Movsziszian-Wl.Schmidt, Dresden 1993.
c4323) 12 exd5 exd5 13 £>b3 (13 ^fl 0-0
14 Af4 d4! 15 £>e5 {15 c4 is better} 15...£>xe5
16 Axe5 dxc3! 17 bxc3 Sd7 18 #c2 Axe5 19
Sxe5 Jta6! + Strikovic-Gheorghiu, New York
Open 1988) 13...d4 14 cxd4 cxd4 15 Af4 0-0
16 £>e5 £>xe5 17 Axe5 Ab7 18 Axb7 Sxb7 19
Scl £rf5 = Damljanovic-Lautier, Pamplona
1992/3.
10...#c711 We2 g512 hxg5 hxg513 £>xg5
13 £tf 1 g4 14£>3h2 #xe5 15 tfdl #h5 16 c4
(16#xg4#xg4 17^ixg4 Aa6!??) 16...Ab7 17
#xg4 Wxg4 18 £>xg4 0-0-0 was fine for Black
in A.Horvath-Kiss, Budapest 1998.
13...Vxe5 (D)
14#xe5£>xe5
W
m§ m
w
iHX
ii
mJSLmwm
3 fiif 4 *tfr'£ ^^
I in
AHA
As in similar positions, 14...Jtxe5!? looks
more accurate, so as not to allow White's next
move. However, I should mention that the main
line is nothing to fear either.
15£>c4
This common trick exploits the slightly
vulnerable a8-hl diagonal. 15 £>df3 £>xf3+ (the
alternative 15...^5c6 also looks quite attractive
for Black) 16 Jtxf3 0-0 was roughly equal but
pleasant for Black in the game Sommerfeld-
Goetz, Bundesliga 1986/7.
15...£>xc4
This would be the choice of most players
because Black risks very little. However, 15...dxc4
16 Jtxa8 cxd3 is not bad either; e.g., 17 £rf3
£>xf3+ 18 Jtxf3 dxc2 and Black has sufficient
play for the exchange, Belkhodja-Pira, Paris
1989.
16 dxc4 Ab7 (D)
W
1
HI
wm A HEP
I wk
ABA
m mm
1A1
I S§ I
m '^bf^wm
17 c3

134
French: Advance and Other Lines
With this move White prepares the
development of his dark-squared bishop and maintains
the tension in the centre. Black has no
problems after 17 cxd5 ±xd5 18 £>e4 Id8 19 c3,
Larrea-M.Socko, Rio de Janeiro 1998, and now
I like 19...f6!.
17...Id8 (D)
Probably more accurate than 17...0-0, after
which White was able to exert some pressure
with 18 £>h3 £>g6 19 £tf4 £>xf4 20 ±xf4
Had8 21 fiadl in Morozevich-Beulen, Hyeres
1992.
18 a4
Now the plan above with 18 ^h3 makes less
sense and can simply be parried by 18...fid7,
but also the more tactical 18...d4! is strong.
The advance of the a-pawn (18 a4) is not that
terrifying though. Even 18...a5 is not a bad
reply. In J.Pedersen-W.Schmidt, Denmark (EU
Ch) 1999, Black played the simple 18...±c6,
with a nice position. Black is ready to take on
c4 or just play ...d4.

9 Wing Gambit and Other Rare Lines
1 e4e6
In this final chapter we shall sum up some of
the rare lines that are at White's disposal. These
include the so-called Wing Gambit, which takes
up the lion's share of these final pages.
Our three main continuations are:
A: 2b3 135
B: 2d4d5 136
C: 2£if3 136
In Line B we focus on White's very rare
options on move 3.
Here is a brief overview of some lesser lines:
a) 2 e5 d6 (or 2...c5 3 f4 £ic6 4 £rf3 f6!? if
one is looking for a more dynamic
continuation) 3 exd6 Axd6 4 d4 £ic6 5 £lf3 £ige7 6
Ad3 0-0 =.
b) 2 f4 d5 3 e5 c5 4 £rf3 £ic6 5 c3 d4 6 d3
£ih6 7 £ibd2 ±e7 8 £ie4 0-0 = Lesny-Kolcak,
Slovakia 2001.
c) 2 g3 d5 3 Ag2 dxe4 4 £ic3 Ad7 5 £ixe4
Jtc6 = Riumin-Rauzer, USSR Ch (Odessa)
1929.
d) 2 c4 d5 3 cxd5 exd5 4 exd5 £lf6 5 Ab5+
(5 £ic3 £ixd5 6 Ac4 £ib4! 7 &f3 £id3+ 8 *f 1
Ae7 9 £M5 £ixcl 10 Ixcl £ic6 11 £ie3 0-0 12
d4 Jtf6 t Geirnaert-Zimmerman, Dieren 2002)
5...£>bd7 6 £ic3 a6 7 Axd7+ Axd7 8 h3 Af5 9
»f3 #d7 10 d3 Ab4 11 Ad2 0-0-0 12 £ige2
£>xd5 is much better for Black, Minzer-Justo,
Buenos Aires 1992.
A)
2 b3 (D)
2...b6
A sort of non-theoretical reply compared to
the more investigated line 2...d5 3 Jtb2 dxe4
(3...£tf6 should also be fine) 4 £>c3 £rf6, and
then:
a) 5 g4!? h6 6 Ag2 Ad7 7 £ixe4 £ixe4 8
Axe4 Ac6 9 «e2 Axe4 10 #xe4 «d5 11
w mm
ll&il w,
A fH A «
#xd5 exd5 12 £ie2 £ic6 13 &f4 0-0-0 14 0-0-0
d4 = Z.Varga-I.Farago, Nagykanizsa 2003.
b) 5 «e2 Ab4 6 0-0-0 «e7 7 *bl (7 £ixe4
±a3 =) 7...Axc3 8 dxc3 Ad7 9 g4 h6 10 Ag2
£\c6 11 Jtxe4 £\xe4 12 #xe4 and now both
12...0-0-0 (Kapnisis-Berelovich, Kavala 2002)
and 12...h5!? (Repkova-Cosma, Pula 2002) are
approximately equal.
3 Ab2 Lhl 4 «e2
The fact that the e4-pawn requires defence
forces White to take some sort of unnatural
measure. Black also equalizes easily against
other moves:
a) 4 d3 d5 5 £id2 &f6 6 e5 &fd7 7 f4 c5 8
£>gf3 Jte7 9 g3 f5 is equal, Johansson-Anton-
sen, Naestved 1991.
b) 4 £ic3 c5 5 £rf3 £ic6 6 d4 cxd4 7 £ixd4
£ixd4 8 #xd4 £rf6 = Bury-J.Watson,
Leominster 1977.
4...a5!?
4...c5 5 £lf3 d6 6 g3 £lf6 7 d3 Ae7 8 Ag2
0-0 9 0-0 £ic6 10 £ibd2 Ic8 11 Ifel d5 with
equality, Gallinnis-Jackelen, 2nd Bundesliga
1998/9.
5 a3 a4 6 b4 c5 7 b5 d6 8 f4 ^d7 9 £if3
£igf6
Black has equalized, Wisnewski-Meister,
Hoeckendorf 2004.

136
French: Advance and Other Lines
B)
2 d4 d5 3 ±d3
3 Jte3?! is the so-called Alapin-Diemer
Gambit where White hopes for speedy
development at the cost of sacrificing a pawn in the
spirit of the Blackmar-Diemer Gambit (1 d4 d5
2 e4). 3...dxe4, and then:
a) 4 f3?! is well met by 4...£>h6!? with the
idea of ...£tf5, while on 4...£\f6 White gets the
type of position he is looking for by 5 fxe4
£>xe4 6 ±d3.
b) 4 £>d2 £>f6 5 c3 (again 5 f3?! is not fully
correct as White's compensation after 5...£M5
6 ffe2 £>xe3 7 #xe3 exf 3 8 £>gxf 3 ±e7 9 ±d3
£>d7 is insufficient) 5...£>c6!? 6 #c2 e5 7 ±b5
±d7 8 ±xc6 ±xc6 9 dxe5 £>g4 10 0-0-0 £>xe3
11 fxe3 »d5 12 ^b3 #xe5 13 £>d4 ±d7 +
Roessler-Galle, Belgium 1997.
3...dxe4
3...c5 is also good. 4 exd5 exd5 is an
Exchange Variation that is equal, but Black can
also play 4...#xd5 when after 5 £>c3 #xd4
(5...#xg2?? 6 ±e4 traps the queen) 6 £>b5 Wd8
7 Jtf4 £te6 White doesn't have full
compensation for the pawn.
4 ±xe4 £tf6 5 ±f3 (D)
Keeping the bishop on the long diagonal
makes more sense than 5 Jtd3.
5...c5 6 £>e2
6 c3 £>c6 7 £>e2 ±e7 8 0-0 0-0 9 lei ±d7 =
Bezgodov-Dolmatov, Elista 2001.
6...£>c6 7 ±e3 cxd4 8 £>xd4 £>e5 9 0-0
£>xf3+10Wxf3#d5
Due to White's lead in development Black
has to step carefully; in Cluvshtein-Hoang
Thanh Trang, Budapest 2003, Black did not
react with enough caution: 10...e5? 11 QSb5 a6 12
Idl ±d7 13 £>d6+ ±xd6 14 Ixd6 Wcl 15
Ixf6! gxf6 16 #xf6 Bg8 17 ±g5 ±c6 18 £>c3
and Black was in serious trouble.
11 We2 a6 12 £>c3 WeS 13 ladl ±d7 14
£>f3#h5
= Bezgodov-Goriachkin, Orsk 2001.
o
2 £>f3 d5 (D)
W
WtkW" WkWk
|U WB,kw§, A.
mm. mm. A m.
ABAB BAH
BSBf'Bl'Bg
Now:
CI: 3£>c3 136
C2: 3e5 137
The former is sometimes called the French
Two Knights, and was recommended by Chris
Baker in his book A Startling Chess Opening
Repertoire. The latter generally signals that
White is intending to play the French Wing
Gambit.
CI)
3 £>c3 (D)
3...d4
Far from Black's only choice. The two main
alternatives are:
a) 3...£rf6 4 e5 £>fd7 5 d4 leads to a line I
covered in The Main Line French: 3 *hc3. For a
refreshment of the main line here is a recent
example: 5...c5 6 dxc5 £>c6 7 ±f4 ±xc5 8 ±d3 f6

Wing Gambit and Other rare Lines
137
9 exf6 £ixf6 10 0-0 0-0 11 £ie5 ±d7 12 #e2
(12 £>xc6 ±xc6 13 ±e5 «e7 14 «e2 £>d7 ?)
12...«e7 13 lael Iae8 14 a3 a6 15 ±g3 £>xe5
16 ±xe5 ±c6 17 b4 (on 17 *hl Black's best
is 17...£id7, when the tactical line 18 1^5 g6
19 ±xg6 £>xe5 20 ±xe8 If5 21 #h3 ±xe8 22
£>xd5 #g7 appears to work fine for Black)
17...±b6 18 b5 axb5 19 £>xb5 ±xb5! 20
±xb5 £>e4 21 ±g3 Id8 22 ±d3 £>xg3 23
hxg3 If6 with a very unpleasant attack against
f2,1.Schneider-Ulybin, Biel 2004.
b) 3...±b4 4 e5 c5 (4...£>e7 is safer and
probably preferable) 5 £ib5!? (5 £ie2 £ic6 6 c3 ±a5
7 d4 cxd4 8 £iexd4 ±c7 9 £ixc6 bxc6 10 ±e3
±b6 11 £id4 thel 12 #g4 0-0 <*> Shivaji-
Sharavdorj, Minneapolis 2005) 5...c4!? (Black's
strategy appeared somewhat misjudged after
5...£>c6 6 b3 £ih6 7 c3 ±a5 8 ±a3 ±b6 9 £id6+
*d7 10 ±e2 f6 11 d4 fxe5 12 dxc5 ±c7 13 ±b5
in Speelman-Cafferty, Hastings 1974/5) 6 c3
±d7?! (6...±c5 7 d3 «b6 8 £>bd4 cxd3 9 ±xd3
£ic6 oo) 7 cxb4 ±xb5 8 a4 ±d7 9 b3 cxb3 10 b5
£ie7 11 ±a3 b2 12 Ibl a6 13 #b3 and White is
better, Brustkern-Vogler, Wiesbaden 1991.
4 Zhel c5 5 c3
5 d3 £ic6 6 g3 ±d6 7 ±g2 £ige7 8 0-0 0-0 9
h4 e5 10 £id2 ±e6 11 f4 f5 ? Guseinov-
Filippov, Moscow 2002.
5...dxc3 6 bxc3 £>f6!? 7 £ig3 £>c6 (D)
8±b5!?
8 d4 cxd4 9 cxd4 ±b4+ 10 ±d2 #a5 11 Ibl
±xd2+ 12 #xd2 4>e7 = Sigurjonsson-Hort,
Hastings 1975/6.
8...±d7 9 0-0 h5! 10 d3 h4 11 £>e2 h3! 12
g3 ±e7 13 ±xc6
13 Ibl!?.
13...±xc6 14 ZheS «c7 15 £>xc6 «xc6 16
f3 0-0-0
Drozdov-Glek, Azov 1996. White has the
more mobile centre but Black has good coun-
terchances as White's king's position is far
from safe.
C2)
3 e5 c5 4 b4 (D)
The Wing Gambit is a dangerous variation
against the French and probably deserves
more respect than it usually gets. To be honest,
I doubt its correctness but in practice White
often gets good chances of launching an
attack.
Black must immediately decide whether to
accept or decline the sacrifice:
4...cxb4

138
FRENCH: ADVANCE AND OTHER LINES
The acceptance of the pawn is the critical
continuation. Black might also consider:
a) 4...c4!? 5 c3 a5 6 b5 £id7 (6...#b6 is best
met by 7 d3!; 6...a4 is also feasible) 7 d4 cxd3 8
Jtxd3 (D) with a choice for Black:
al) 8...»c7 9 «e2 £ie7 10 0-0 £ig6 11 lei
±e7 12 h4 ±xh4 (12...£ixh4 13 £ixh4 ±xh4
looks more critical; e.g., 14 #g4 £>xe5 15 #xg7
±xf2+ 16 4>xf2 £ixd3+ 17 *fl Bf8 18 ±h6
«c5 19 Be3 ±d7 20 «xf8+ Wxf8 21 ±xf8
4>xf8 shouldn't be bad for Black) 13 £ixh4
£ixh4 14 ±a3 #d8 15 #h5 £ig6 <*> Morten-
sen-J.Hansen, Gausdal Z 1987.
a2) 8...f6!? 9 #c2 £ixe5 10 £ixe5 fxe5 11
±xh7 ±d6 12 ±xg8 Ixg8 13 c4 (13 #h7!?)
13...«c7 14 ±a3 ±b4+!? 15 ±xb4 axb4 16
£id2 e4 17 «b2 Bh8 18 g3 e3 19 fxe3 Ba3
with counterplay, Nayer-Moskalenko, Moscow
1995.
b) 4...d4!? 5 bxc5 ±xc5 6 ±a3 (this is
considered critical) 6...#a5 7 ±xc5 #xc5 8 c3
£ic6 9 cxd4 (9 #b3 might be better) 9...£>xd4
10 #a4+ Jtd7! (a strong improvement over
10...£ic6 11 £ic3 «a5 12 «g4 *f8 13 Wf4
£ige7 14 ±d3 £ig6 15 ±xg6 hxg6 16 0-0 with
a big advantage for White, Boersma-Korchnoi,
Leeuwarden 1977) 11 #xd4 #cl+ 12 4>e2
±b5+ 13 4>e3 (13 d3 Bd8 14 £ibd2 «a3 looks
too dangerous for White) 13... Jbcf 1 + Deman-
Haberkorn, Antwerp 1999.
c) 4...b6 (a good practical decision: Black
keeps d4 under control by maintaining his pawn
on c5) 5 c3 (D) (5 d4!? cxd4 6 a3 could be worth
looking into) and then:
cl) 5...a5!? 6 b5 c4 is similar to line 'a' but
Black has already 'wasted' a move on ...b6.
c2) 5...Wd7 (aiming to exchange the light-
squared bishops with a later ...Jta6) 6 a3!? d4
(6...±a6 7 d3!? is White's idea) 7 cxd4! cxb4 8
d5! #xd5 (8...exd5 9 d4 is also better for White
according to Mortensen) 9 axb4 Jtxb4 10 #a4+
£ic6 11 ±b5 ±d7 12 0-0 £ige7, Mortensen-
Bjerring, Denmark 1986, and now 13 Bdl!
intending £ic3 is very good for White.
5 a3 (D)
White can avoid Line C22 below by playing 5
d4 first but then 5...±d7! 6 a3 (6 ±d3 could be
considered but then when White plays a3 and
Black replies ...^c6, a black capture ...^xb4
will come with an attack on the bishop) 6...#a5!
7 Ad3 (7 ±d2!?) 7...±b5! and Black was
already better in Day-Hiibner, Jerusalem jr Wch
Now there are two main lines for Black:

Wing Gambit and Other Rare Lines
139
C21: 5...&C6 139
C22: 5...d4!? 140
Black can of course also take on a3: 5...bxa3
6 d4 ^c6 7 c3 and now Black's main choice
consists of deciding whether to attack White's
centre with ...f6, or aim for a blockade with
...f5:
a) 7...±d7 8 ±d3 f5 - Z../5 8 JLd3 JLd7.
b) 7...f6 8 ±d3 fxe5 9 dxe5 (9 £ixe5!? is
another idea) 9...#c7 10 #e2 £>h6 11 0-0 (11
±xh6!?) Il...£tf7 12lel g6 13 £>xa3 ±c5 14
£>b5 #d8 15 Ia4!? with compensation for
White, S.Bucker-Poldauf, German Ch (Bad
Neuenahr) 1991.
c) 7...f5 8 ±d3 £>h6!? (8...±d7 9 g4!? £>ge7
10 gxf5 £>xf5 11 £>g5 £>ce7 12 £>xa3 h6 13
£>h3 «c7 14 £>b5 ±xb5 15 ±xb5+ 4>f7 16
Jtd2 with unclear play, Mortensen-Karlsson,
Copenhagen 1985) 9 Igl (9 ±xh6 gxh6 10
£>xa3 ±xa3 11 Ixa3 #e7 12 lal 0-0 + Bosch-
Peelen, Amstelveen 1994) 9...±e7 10 ±xh6
gxh6 11 g4 fxg4 12 Ixg4 «a5 13 «b3 «b6 14
«c2 «b2 15 «xb2 axb2 16 Ia2 ±d7 17 Ixb2
0-0-0 + Djurhuus-Hellsten, Oslo 1997.
C21)
5...£>c6 6 axb4 Axb4
6...£>xb4 7 c3 £>c6 - 6...±xb4 7 c3 &.J8.
7 c3 (D)
7...±e7
This is the most common retreat but two
other bishop moves are also seen:
a) 7...Jtf8 has been played by Karpov but
"doesn't make any sense" according to
Harding. Indeed, he has a point, since if White
plays 8 d4 (the most logical), Black could have
had the same position with a pawn on a3, if he
had played 5...bxa3.
b) 7.. JLa5 (this contains a large degree of
logic: the bishop isn't going to obstruct the
knight but is also removed as a potential
defender on the kingside) 8 d4 and now:
bl) 8...f6 9 ±d3!? fxe5 10 dxe5 £>ge7 11
0-0 #c7 12 £>a3 a6 13 c4 d4 14 c5 was
promising for White in Villumsen-Ryan, corr. 1991.
b2) 8...£>ge7 9 ±d3 h6 10 0-0 (10 h4! -
Harding) 10...a6 11 ±a3 ±d7 12 g3 b5 13 ±c5
±b6 14 ±xb6 «xb6 15 £>bd2 £>a5 16 «e2 <*>
Westerinen - Brinck-Claussen, Esbjerg 1978.
8 d4 (D)
To compensate for the lost pawn, White has
open files on the queenside and a strong centre.
It is on the kingside, though, that White usually
bases his chances with the proud e-pawn taking
a lot of space. For the moment, Black also has
to think about how to develop his king's knight,
since after ...£tfi6 White can double his pawns
on the h-file.
8...f6
This is regarded as the strongest but Black
can also wait with kingside development and
instead seek activity on the queenside
immediately with 8...±d7!? 9 ±d3 a6 intending ...b5
and ...£te5-c4. However, Black cannot leave his
kingside untouched forever, and thus White
should simply counter Black's plan by building

140
French: Advance and Other Lines
up kingside pressure. Best seems 10 h4!?, and
now:
a) 10...£>a5 11 £>g5 £>h6 (ll...h6 12 £>xf7
*xf7 13 Wf3+ £>f6 14 exf6 ±xf6 15 Af4 with
an attack) 12 £>xh7 £>f5 13 Axf5 exf5 14 Wh5
g6 15 ®h6 Af8 16 Wg5 Ae7 (16...Sxh7 17
Wxd8+ Sxd8 18 Sxa5 •) 17 Wh6 =.
b) 10...h6 11 Sh3 £>a7 12 Sg3 *f8 13 h5
Ab5 14 Ac2 Ae8 15 £>gl f5 16 Sh3 <*>.
9 ±d3 fxe5 (D)
W
10 dxe5
10 £>xe5 £>f6 11 £g5 £>xe5 12 dxe5 £>e4 13
±xel ®xe7 14 ±xe4 dxe4 15 Wd4 0-0 16
Wxe4 (16 Sxa7 Sxa7 17 Wxa7 Wg5 was better
for Black in Buturin-Glek, Belgorod 1989)
16...Ad7! 17 We3 (on 17 Wxb7, 17...Wc5! is
strong) 17...£b5 18 £>d2 Wc7 19 f4 Sad8 +
Reinderman-Glek, Groningen 1992.
10...&H6
The knight has to go this way. It is possible
to make room for it on e7, say by 10...JLc5, but
Black then removes a vital defender of the
kingside, viz. the dark-squared bishop.
Alternatively, Black can first tie White's queen to the
defence of the e-pawn with 10...Wc7 11 ife2
(11 0-0 £>xe5 12 £>xe5 Wxe5 is not a healthy
pawn sacrifice as shown in Hamed-Garma,
Novi Sad OL 1990 when, after 13 ±b5+ *f7!,
Black's king could be moved safely to the king-
side after ...&f6 and ...Sf8/d8) and only now
1 l...£ih6. This appears to be to Black's
advantage but isn't entirely clear. White now
continues 12 £.xh6 gxh6 13 0-0 (13 We3 is met by
13...Ag5) 13...0-0 14 We3 Sf7, Lampen-Rud,
Gausdal 2000, and now 15 £\a3! ±d7 16 £\b5
WbS 17 c4! looks like good compensation for
White.
11 ±xh6 gxh6 (D)
W
12Wd2
Instead after 120-0 0-0 13 Sa4 Sf7 14 2g4+
fig7 15 Sg3 Wf8!? Black defended
successfully in Salmensuu-Crawley, Tanta City 2000.
Maybe the immediate 12 Sa4 should be
considered.
U...±g5
A fairly logical move since Black now gets a
tempo by attacking the queen but the bishop is
not that well placed. 12...Wc7 is critical, though
White may claim some compensation after 13
Wxh6 £>xe5 14 Wh5+ £>f7 15 ±b5+ *f8 16
0-0. 12.. Jtf8!? was suggested by Hillarp-Pers-
son, intending .. Jtg7 and ...0-0. The bishop is
going to be very useful on g7, attacking e5.
13 We2 Wc7 14 0-0 0-0 15 £>a3 Wg7 16
£>b5 ±dl
16...±f4!? 17 Sfel ±dl could also be
considered.
17 £>d6 ±e8 18 £>xg5 hxg5 19 Sfel
White has reasonable compensation for the
pawn, Salmensuu-S.Pedersen, Torshavn 2000
and Salmensuu-Manninen, Finnish Cht 2000.
C22)
5...d4!? (D)
An ambitious attempt to prevent White from
strengthening his centre with d4. But every
pawn move leaves weaknesses behind and the

Wing Gambit and Other Rare Lines
141
w
VMikW
w*y,
WB k HP k
« ym kW/ 9
v Wk, Wm*k Wk, mm
mm, mm
text-move leaves White with better control of
the light squares and the d4-pawn is itself a
potential weaknesses if White can find time to
gang up an attack against it.
6.4b2
Probably best. White hopes to benefit from
being able to take on b4 only later, and if Black
captures on a3 White develops his knight.
Instead, 6 axb4 ±xb4 7 ±b2 (7 ±a3 ±xa3 8
£\xa3 £\e7 is nothing for White) 7..Jtc5 (this
may be more accurate than 7...£>c6, after which
White usually tries 8 ±b5 ±c5 9 We2 and can
attempt to open the centre with c3 thereafter) 8
±b5+ (8 Ad3 £>c6 9 0-0 £>ge7 10 Sel £>g6
was good for Black in Ciuffoletti-Vallejo Pons,
Saint Vincent 1999) 8...Ad7! 9 ±xd4 ±xd4 10
£>xd4 Wb6 11 ±xd7+ £>xd7 12 £rf3 £>e7 gives
Black a comfortable position.
6...£>c6 7 ±b5 ±d7 8 0-0 (D)
More accurate than 8 We2 a6!, and now:
a) 9 Ad3 Ac5 (9...Wb6!?) 10 axb4 £>xb4
11 ±xd4 ±xd4 12 £>xd4 Wb6 + Dannevig-
Djurhuus, Norway 1990.
b) 9 ±xc6 ±xc6 10 £>xd4 ±xg2 11 Sgl
bxa3(ll...Ad512axb4*h413c4Wxh214*fl
Wh3+ 15 *el Wh2 = Re-Miana, Rosario 1992)
12 £>xa3 Ad5 13 £>ab5! axb5 (13...Sc8!?) 14
Wxb5+±c6 15 £>xc62xal+ 16 Axal bxc6 17
Wxc6+ = Djurhuus.
8...£>ge7
Alternatives:
a) 8...£>xe5? 9 £>xe5 ±xb5 10 £>xf7! <&xf7
11 Wh5+ g6 12 Wxb5 with the better game for
White, S.Bucker-Uhlmann, German Ch (Bad
Neuenahr) 1991.
b) 8...£>h6 9 ±xc6 ±xc6 10 £>xd4 ±d5 11
axb4 ±xb4 (ll...Wg5!?) 12 c4 ±c6 13 £>xc6
bxc6 14 Wa4 Wb6 15 Aa3 (15 Ad4 Wb7 16
Ac3 ±xc3 17 £>xc3 0-0) 15...±xa3 16 Wxa3
c5 17 Wa4+ <&e7 18 £>c3 Shd8 19 £>e4 ±
Bisby-Kustanovich, Witley 1999.
c) 8...Wb6!? 9 ±xc6 ±xc6 10 £>xd4 bxa3
ll±c3oo.
9 We2 £>g6 10 We4 a6 11 ±d3 Wc7 12 Sel
Ac5 13 h4 h5 14 g3 Sd8 15 We2 (D)
m
gkmxwxm
m '//////ft **<v '//////// m V////?/ *A^ '//////A
White has compensation, Gajewski-Szoen,
Zakopane 2001.

Index of Variations
Chapter Guide
1 e4 e6
2 (14
2 We2 - Chapter 8
2 d3 - Chapter 8
2 £>f 3 - Chapter 9
Other moves - Chapter 9
2 ... d5
3 e5
3 exd5 - Chapter 7
Other moves - Chapter 9
3 ... c5
Other moves - Chapter 6
4 c3
Other moves - Chapter 6
4 ... £>c6
Other moves - Chapter 6
5 £>f3
Other moves - Chapter 6
5 ... #b6
5..Adl- Chapter 4
5...£\ge7- Chapter 5
5...£\h6- Chapter 5
After 5...Wb6:
6 a3 - Chapter 1
6 ±e2 - Chapter 2
6 ±d3 - Chapter 3
1: Advance: 5...1b6 6a3
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 ^c6 5 £>f3 #b6
6a3 73
A: 6...f6!? 7J
B: 6...a5 74
C: 6...£>h6 161 b4 cxd4 8 cxd4 £>f5 77
Cl:9Ae3iS
C2: 9 Ab2 27
C21:9...Ad7 27
C22: 9...Ae7 25
D:6...c4 2S7£>bd2 2<3
Dl:7...f6 29
D2: 7...£>a5 37
D21:8Ae2J7
D22: 8 g3 33
E: 6...Ad7 56 7 b4 cxd4 8 cxd4 Ic8 36
El: 9 Ab2 37
E2: 9 Ae3 39
2: 6 ±e2
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 ^c6 5 £>f3 #b6
6 Ae2 42
A: 6...£>h6!? 42
B: 6...cxd4 441 cxd4 £>h6 45
Bl: 8 b3 45
B2: 8 £>c3 46
3: 6 ±d3 and the Milner-Barry
Gambit
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 £>c6 5 £>f3 #b6
6 ±d3 45 6...cxd4 7 cxd4 Ml 8 0-0 £>xd4 48
A: 9 £>g5 49
B: 9 £>xd4 50 9...#xd4 10 £>c3 50
Bl: 10...a6 57
B2: 10...#xe5 52
4: 5...±d7
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 ^c6 5 £tf3 i.d7
54
A: 6 dxc5 55
B: 6 a3 5(5
Bl:6...Sc8 5(5
B2: 6...£>ge7 55
B3:6...c4 59
B4: 6..i6!? 61
C: 6 Ae2 63
Cl:6...Wb6«
C2:6...f6!? 65 7 0-0 65

Index of Variations
143
C21:7...Wb6 66
C22: 7...fxe5 68
C3: 6...Sc8 70
C4: 6...£>ge7 77
C41: 7 0-0 72
C42: 7 £>a3 74 7...cxd4 8 cxd4 £>f5 9 £>c2 75
C421:9...Ae7 76
C422: 9...£>b4 77
C423: 9...ffb6 79
5: 5...<£>ge7 and 5...<£>h6
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 £>c6 5 £>f3
A: 5...£>ge7 84
B:5...£>h6 87
Bl:6dxc5S7
B2: 6 £d3 89
6: Various Deviations
1 e4 e6 2 d4 d5 3 e5 97
A: 3...b6 97
B: 3...£>e7 92
C: 3...c5 93
C\: 4 Wg4 93
C2: 4 dxc5 94
C3: 4 £>f3 95
C4: 4 c3 96
C41: 4...#b6 96
C42: 4...£>c6 700
7: Exchange Variation
1 e4 e6 2 d4 d5 3 exd5 exd5 702
A: 4 £>c3 702
B:4.£.d3 703
C: 4 c4 704
D: 4 £>f3 706
Dl:4...±g4 706
D2: 4...£>c6 708
D3: 4...£>f6 709
D4: 4...Ad6 777
8: King's Indian Attack
1 e4e6
A:2#e2 772 2...c5 772
Al:3f4 774
A2:3g3 776
B:2d3 777 2...d5 777
Bl:3«e2 777
Bll:3...dxe4 777
B12:3...c5 720
B2: 3 £>d2 723 3...c5 4 £>gf3 £>c6 5 g3 g6 6
Ag2 Ag7 7 0-0 723
B21:7...d4!?723
B22: 7...£rf6 725
B23: 7...£>ge7 726 8 Sel b6 727
B231:9c3 727
B232: 9 exd5 729
B233: 9 h4 737
9: Wing Gambit and Other Rare
Lines
1 e4 e6 735
A: 2 b3 735
B:2d4d5 736
C: 2 £rf3 736 2...d5 736
Cl:3£>c3 736
C2: 3 e5 737 3...c5 4 b4 cxb4 5 a3 738
C21:5...£>c6 739
C22: 5...d4!? 740

In the third and final volume of his series on the French Defence, Steffen
Pedersen covers all of White's alternatives to 3 ?hc3 and 3 £}d2. These include
the aggressive Advance, the quiet Exchange, and side-variations such as the Two
Knights, King's Indian Attack and the Wing Gambit. In these lines, it is White who
determines the central structure and sets the pace of the game.
With the Advance Variation, White establishes a space advantage, and will often
seek to exploit this by creating attacking chances on the kingside, or a bind on the
queenside. Both sides have a variety of systems at their disposal, and the player
who is better acquainted with the intricacies of these lines will tend to come out on
top. It is a perennial favourite among players with a direct aggressive style: Shirov
has played the Advance many times in recent years, while it has been employed
by Sveshnikov and Kupreichik throughout their careers.
The Exchange Variation leads to open and generally symmetrical positions.
Pedersen carefully examines White's attempts to seize the initiative, and also
recommends ways for Black to create play if White adopts a more stolid
approach.
Steffen Pedersen is an international master from Denmark who scored a
grandmaster norm when he was 16, and played in several World and European
junior championships. He was joint winner of the Hampstead Grandmaster
tournament in 1998, achieved his second grandmaster norm at the Aars get2net
Cup in 1999, and won the Danish Championship in 2004. Pedersen has written
several highly-regarded opening books, with subjects including the Dutch
Defence, Sicilian Scheveningen, Benko Gambit and the Semi-Slav.
Other titles from Gambit Publications include:
The Main Line French: 3 ^c3
Steffen Pedersen
The French: Tarrasch Variation
Steffen Pedersen
How to Beat 1 d4
James Rizzitano
Win with the London System
Sverre Johnsen and Vlatko Kovacevic
Play the Sicilian Dragon
Edward Dearing
Gambit Publications Ltd is:
Managing Director: Murray Chandler GM
Chess Director: Dr John Nunn GM
Editorial Director: Graham Burgess FM
For further information about Gambit Pub
send an e-mail to: info@gambitbooks.cor
http://www.gambitbooks.com
The English Attack
Tapani Sammalvuo
Mastering the Najdorf
Julen Arizmendi and Javier Moreno
A Complete Guide to the Grivas Sicilian
Efstratios Grivas
Play the 4 f3 Nimzo-lndian
Yuri Yakovich
Play the Classical Dutch
Simon Williams
£14.99 $24.95
ISBN 1 TDMbOD MD 1
9»7 8 1 904ll600404»>