ChessShelf

King's Indian Defense: Petrosian Variation

Named variation

In the most common King's Indian move order, White castles first — 7.O-O — before deciding what to do about the center, which gives Black an extra developing move as well. The Petrosian Variation changes that order: after 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 O-O 6.Be2 e5, White plays 7.d5!? immediately, locking the center a move before castling rather than after. That single reordering, named for the deeply prophylactic style of Tigran Petrosian, is a deliberate attempt to take away one of Black's central options before Black has had the extra tempo to prepare around it.

What locking the center early is actually preventing

The standard King's Indian counterplay — a kingside pawn storm with ...f5, ...f4, and eventually a direct attack on White's king — depends on Black having time and support to prepare it properly. By locking the center immediately with d5, White denies Black some of the flexibility that extra tempo would otherwise provide, and typically follows up with Bg5, pinning or pressuring the f6 knight and making ...f5 harder to prepare cleanly. This is prophylaxis in the most literal sense: not reacting to Black's kingside plan once it starts, but interfering with the setup that plan depends on before it has a chance to develop.

The other half of the plan: queenside expansion

Locking the kingside down is only half the point — White also uses the resulting closed center to expand calmly on the queenside with moves like b4, a4, and c5 at the right moment, converting the prophylactic restraint on one wing into concrete progress on the other. This two-part structure (restrain Black's plan, then advance your own on the opposite side) is the entire strategic identity of the Petrosian Variation, and it's why the line carries Petrosian's name specifically — this alternating restrain-then-expand approach was a hallmark of how he handled closed positions generally, not just in this one opening.

How this differs from simply 'the King's Indian with an early d5'

Locking the center with d5 happens in many King's Indian lines eventually, so the timing, not the pawn move itself, is what actually earns this the Petrosian name. What makes it specifically the Petrosian Variation is doing it before castling, denying Black the extra preparatory move a later d5 would concede. A White player who reaches a structurally similar locked position several moves later, after castling and other development, has not really played the Petrosian idea — the early timing is the entire point, not an incidental detail of how the position happened to arise.

There's also a piece-placement detail worth knowing on Black's side of the ledger: with the center locked, a knight that would normally sit comfortably on f6 can become a target for White's Bg5, and rerouting it — often via e8 or d7 toward c5 or f4 later — takes real time that a less restrained White setup wouldn't force. Black's minor pieces generally need a more deliberate reorganization plan in the Petrosian structure than in the faster-moving Classical Variation, since the usual routes to active squares are partially blocked by the same locked center White worked to establish early.

The order the plan actually needs

d5 first, to deny the tempo. Bg5 next, to make ...f5 harder to prepare cleanly. Queenside expansion after that, once the kingside restraint is in place. Playing these out of order — expanding on the queenside before the kingside restraint is actually established — gives Black time to prepare the standard counterplay the whole plan exists to prevent.

The recurring White mistake is losing patience with the prophylactic half of the plan — locking the center and then rushing straight to queenside expansion without first making Black's kingside plan genuinely difficult with Bg5 and careful piece placement. Skipping that step lets Black prepare ...f5 anyway, and the whole point of playing d5 a move early gets wasted. The matching Black mistake is trying to force the standard kingside storm regardless of the specific restraint White has set up, rather than adjusting preparation — Black often needs an extra move like ...Nh5 or ...Ne8 to properly support ...f5 here that wouldn't be necessary against a less prophylactic White setup, and skipping that adjustment just walks into exactly the difficulty White engineered.

For the maximally ambitious opposite approach within the same opening, see the Four Pawns Attack, and for the restrained, mirroring approach that avoids the center fight altogether, the Fianchetto Variation.

Frequently Asked Questions

Is the Petrosian Variation considered a quiet, low-risk choice for White?

Relatively, yes — compared to the Four Pawns Attack it involves far less overextension risk, since White isn't racing against an incomplete development. It still requires precise handling of the restrain-then-expand plan, but the underlying structure is considerably more stable.

Does locking the center with an early d5 give up any winning chances for White?

It trades some dynamic potential for structural safety, similar to other prophylactic approaches across chess generally. The compensation is a lasting, hard-to-dislodge space advantage and a genuine strategic plan on the queenside, rather than a race that could go either way.

Can Black avoid this specific structure by delaying ...e5?

Delaying or varying the central pawn break is a real option and does sidestep this exact move order, though it usually just leads to a different King's Indian structure with its own set of challenges rather than avoiding the opening's core strategic questions altogether.