Queen's Gambit Accepted: Rosenthal Variation – Nc3 (Black's Guide)

The move 4…a6 looks modest, but it serves several purposes in one go. First, it prepares …b5, letting you grab space on the queenside and try to hold onto the extra pawn on c4. Second, it prevents White's pieces — especially knights and bishops — from using the b5 square to pressure your position. Finally, it keeps the game flexible: you haven't committed your extra pawn's fate yet. If White plays quietly, you can develop and later decide whether to keep the pawn or return it for a strong central setup. It's a solid, modern approach to the Queen's Gambit Accepted that avoids the heavily analysed main lines.

Stockfish's top recommendation for White is a4 — a direct attempt to stop your …b5 plan. The suggested continuation runs a4 Nc6 e3 Na5. Notice that White spends a move on a4, then you quickly challenge the centre with …Nc6 and …Na5, targeting the bishop on c1 and keeping an eye on the c4 pawn. The engine evaluates this at +0.04, essentially perfectly equal. In practice, White scores 57.6% from this line — a high number — but that likely reflects skill differences rather than the opening itself. Over 43,075 games, White's results are boosted by stronger players choosing a4, but the objective truth is that the position is balanced.

White's most common choice is 5.e4, played over 61,000 times. It's natural: White grabs the centre and develops. But here White scores just 49.9% — actually below average — suggesting that Black is doing fine. After 5.e4, you can continue …b5, …Bb7, and …Nd7, or even …c5 to strike back immediately. The second most popular is 5.a4, which we covered above. Then comes 5.e3 (24,088 games, White 52.0%), a quieter development that often transposes into a normal Queen's Gambit Accepted with an extra …a6 thrown in. After 5.e3, you can play …Nf6, …Bd6 or …Be7, and aim for …0-0 before deciding on your queenside expansion.

Two moves by White stand out as clear errors in this position. First, 5.Bf4 is an inaccuracy that costs White about half a pawn — the engine says 5.e4 was better. In that line, you can calmly develop with …Nf6 or even …b5, knowing White's bishop on f4 doesn't threaten much. Second, and much worse for White, is 5.Qa4+. The stats show this scores just 42.4% for White across 3,633 games — the worst of any common move. The engine calls it a mistake, losing roughly 1.5 pawns. After 5.Qa4+, you can block with …Bd7, and if 6.Qxc4 b5, you win a tempo while White's queen has to move again. White's best was also 5.e4 here, so when your opponent plays Qa4+, you're already doing well.