The Scotch Gambit: a complete guide to 4.Bc4, the open centre and the attack on f7

Short answer

The Scotch Gambit (1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Bc4) trades a central pawn for immediate development, an open centre and pressure on f7. This guide covers the modern main line with 5.c3 Nf6 6.e5, why the Möller Attack stopped being an elite choice, what to do against Black's two ways of giving the pawn back, and the Max Lange Attack against 4...Nf6.

The essentials in a few lines

• The gambit is not about the pawn: it is about arriving first. After 4.Bc4 White has an active bishop, a half-open e-file and a knight ready to go, while Black has an extra pawn on d4 that will cost time to hold.
• The modern line is 5.c3 Nf6 6.e5, not the Möller recapture 6.cxd4: the engine keeps a small white edge in the first and evaluates the second as dead level.
• The Möller Attack is playable and still beautiful, but it ends in equality against accurate defence — playing it today is a bet on your opponent's memory, not on the evaluation.
• Against 5...dxc3? the strong move is not the natural recapture: 6.Bxf7+! wins the bishop on c5 with a queen check and leaves White close to a pawn ahead on the evaluation.
• Against 4...Nf6 the Max Lange Attack (5.O-O Bc5 6.e5 d5 7.exf6 dxc4 8.Re1+) scores heavily in practice; the engine prefers 9.fxg7 to 9.Ng5, and it is worth knowing that before betting the game on the line.

What the Scotch Gambit offers

The trade is the one every opening gambit makes: permanent material for temporary time. What gives it substance here is the structure. With a black pawn on d4 and a white pawn on e4, the centre opens asymmetrically, and White's pieces have natural squares — c3 for the knight or the pawn, e1 for the rook, and the bishop on c4 aiming at the square Black's king has not finished protecting.

There is a second, practical reason for its popularity at club level: it sidesteps the Ruy Lopez. Players who answer 1.e4 with 1...e5 usually prepare the Spanish carefully, spend far less time on the Scotch and almost none on the gambit — which takes the game out of their repertoire on move four.

The price is known and worth stating up front: against accurate defence the compensation gives equality, not an advantage. The gambit pays while the opponent does not know the line, and the positions below show exactly where that border runs.

The modern line: 5.c3 Nf6 6.e5

The difference between 6.e5 and 6.cxd4 is one of move order, and it is enormous. After 6.cxd4 Black gets 6...Bb4+, a developing check that forces White into an uncomfortable decision on move seven. After 6.e5 that check never happens: the knight on f6 has to move first, and the most natural defence, 6...d5, hands over the e5 square and the tempo White wanted.

After 6...d5 7.Bb5 Ne4 8.cxd4 Bb6 the position has a clear profile. White has the pawn on e5 cutting the board in half, the c-file open and the pin on c6 available; Black has a well-placed knight on e4 and a bishop pointing at the centre. It is a manoeuvring game, not a lightning attack — and that is what the gambit's main line has become.

For anyone playing 1.e4 who wants a usable repertoire, this is the version worth memorising. It is six moves, always the same, and understanding the squares — e5 for the pawn, b5 for the bishop, d4 recovered on move eight — solves far more games than any long trap line.

The Möller Attack and why it faded

The sequence is one of the most beautiful in nineteenth-century chess and remains so. White lets the knight on c3 be taken, answers with 9.d5 shutting the black bishop's diagonal and recovers the material with the rook on e4, reaching a position where almost every piece is in play on both sides.

The problem is the verdict. After 9...Bf6 10.Re1 Ne7 11.Rxe4 d6 12.Bg5 the evaluation is exactly zero: Black has given back what needed to be given back and the position is balanced. That is not a refutation, it is the expected result of two accurate defences — and it is why the line all but vanished from top-level chess without ever being disproved.

There is one move-order detail that decides plenty of club games: on move seven the engine prefers 7.Bd2 to 7.Nc3. It is the dull move, trading bishops and keeping the structure, and it holds more practical advantage than the sacrifice. Anyone playing the Möller should know they are choosing the more entertaining line, not the better one.

Giving the pawn back: 5...dxc3 and 5...d3

The capture 5...dxc3 is the most common reply from players meeting the gambit for the first time, and it is a mistake. The engine does not recommend the natural recapture 6.Nxc3 — which restores the balance and actually leaves Black slightly better — but 6.Bxf7+!, exploiting the fact that the bishop on c5 and the king on e8 sit on the same diagonal as the coming queen on d5.

After 6...Kxf7 7.Qd5+ Kf8 8.Qxc5+ White has won the bishop, recovers the pawn on c3 next move and is close to a pawn ahead on the evaluation, against a black king that has lost the right to castle. That is the best defence; the alternatives, such as 7...Ke8 8.Qh5+ g6, make things considerably worse for Black.

The other way of returning the pawn is different in spirit: 5...d3 does not want material, it wants to clog the position and hand the pawn back on the most annoying square, taking White out of the open game. The answer is to take the tempo before the pawn — 6.b4! Bb6 7.a4 a6 and only then 8.Bxd3, with queenside space and an intact structure.

Against 4...Nf6: the Max Lange Attack

The sequence is 6...d5 7.exf6 dxc4 8.Re1+ Be6 9.Ng5 Qd5 10.Nc3 Qf5 11.Nce4, with White attacking f7 and the bishop on e6 at the same time and Black holding an extra pawn he cannot use. After 11...Bf8 12.Nxf7 the engine's evaluation sits around +0.7 — a consistent edge, in a position where one error ends the game immediately.

Honesty about the line cuts both ways. For White: on move nine the engine prefers 9.fxg7 and rates 9.Ng5 as slightly inferior. For Black: after 9.Ng5 the defence is 9...g6, not 9...Qd5 — which is precisely the natural move, and explains why the line scores as heavily as it does.

That makes the Max Lange a choice about time control. In rapid or blitz, forcing your opponent to find 9...g6 on the clock is an excellent deal. In a classical game against someone prepared, the line remains playable, but the return is far smaller than its reputation suggests.

How to learn the gambit in two weeks

In week one, play only the modern line. After 4...Bc5 5.c3 Nf6 6.e5, let Black choose and solve the positions at the board, with no reference during the game. The point is not to get everything right: it is to find out which black replies you cannot answer, and there are always fewer of them than it feels like.

In week two, add 4...Nf6 and the Max Lange, and revise the two pawn-return lines. It is worth setting up the positions from this guide's diagrams and playing them out against the computer from move ten — the hard part of the gambit is not the opening, it is the middlegame it produces, and that is not learned by memorising moves.

After that, what remains is measurement. Score by variation, and the average move at which the initiative runs out, tell you within twenty games whether the gambit is paying for you or merely handing a pawn to opponents who already know the defence.

Frequently asked questions

Openings mentioned in this guide

• Scotch Gambit (C44)
• Scotch Game (C45)
• Italian Game (Giuoco Piano) (C50)
• Danish Gambit (C21)
• Ruy Lopez (Spanish Opening) (C60)