1. e4 e5 2. Nf3 Nc6 3. Bc4 f5
The Rousseau Gambit is a counterattacking chess opening for Black against the Italian Game: 1. e4 e5 2. Nf3 Nc6 3. Bc4 f5!?. Instead of the quiet main lines with 3... Bc5 or 3... Nf6, Black immediately offers the f-pawn — essentially playing a King’s Gambit with colors reversed, a tempo down. Objectively it is risky, but it drags Italian players out of everything they know on move three, and the positions reward whoever attacks better, not whoever memorized more.
The gambit arises only through the Italian move order: 1. e4 e5 2. Nf3 Nc6 3. Bc4, and now 3... f5 defines the opening. The early 2... Nc6 matters — without it, an immediate 2... f5 would be the Latvian Gambit, a different and even wilder animal. From the starting position White faces an immediate decision: take the pawn with 4. exf5, when 4... e4 chases the f3-knight all the way back to g1, or decline with 4. d3, keeping the strong point on e4 and asking Black to prove compensation. Side tries like 4. d4 and 4. Nc3 lead to open, concrete play where both sides need to know what they are doing.
1. e4 e5 2. Nf3 Nc6 3. Bc4 f5 4. exf5 e4 5. Ng1 d5 6. Bb5 Qg5 7. g3 Qxf5
White takes the pawn and Black shows the point: 4... e4 hits the knight, and after the humiliating retreat to g1 White has effectively undeveloped. Black builds a big center with ...d5, and ...Qg5 attacks both g2 and the f5-pawn at once, regaining the material with a clear lead in development. White’s kingside pieces are still in the box while Black is ready to castle long and attack.
1. e4 e5 2. Nf3 Nc6 3. Bc4 f5 4. exf5 e4 5. Qe2 Qe7 6. Ng1 Nf6
White pins the e4-pawn with the queen instead of retreating immediately. Black calmly mirrors with 5... Qe7, defending e4, and the knight goes back to g1 anyway. Black develops with ...Nf6 and ...Nd4, harassing the early queen, while the e4-pawn cramps White’s whole kingside. White’s extra pawn on f5 is hard to keep and harder to enjoy.
1. e4 e5 2. Nf3 Nc6 3. Bc4 f5 4. d3 Bc5 5. Nc3 f4 6. O-O d6 7. a3 Bg4
Declining with 4. d3 is the theory-approved answer: White keeps the e4 strong point and refuses to open lines. Black switches to a reversed King’s Gambit Declined plan — ...Bc5, then ...f4 grabbing kingside space, ...d6, and ...Bg4 pinning the f3-knight. The game becomes a slow-burn attacking race in which Black’s space on the kingside gives real practical chances despite the nominal tempo deficit.
1. e4 e5 2. Nf3 Nc6 3. Bc4 f5 4. d3 Bc5 5. Ng5 f4 6. Nf7 Qh4 7. O-O Nf6 8. Nxh8 Ng4 9. h3 Nxf2 10. Rxf2 Qxf2+ 11. Kh1 f3 12. Qxf3 Qg1#
The knight hop to g5, eyeing the classic f7 fork, is exactly the natural move the Rousseau is designed to punish. Black ignores the rook: 5... f4 and 6... Qh4 launch every piece at White’s king while the knight wanders to h8. After ...Nf6–g4 and the sacrifice on f2, the attack crashes through by force — the line above ends in checkmate while White is still materially ahead, the wayward knight stranded in the corner on h8.
1. e4 e5 2. Nf3 Nc6 3. Bc4 f5 4. d4 exd4 5. e5 d5 6. exd6 Qxd6 7. O-O Nf6 8. Re1+ Be7
White answers the flank strike with a central one. After 4... exd4 5. e5, Black should not panic: 5... d5 blocks the bishop’s diagonal, and after the en passant capture and ...Qxd6 the position opens up with roughly balanced chances. Black needs to develop quickly and meet the e-file pressure with ...Be7 and short castling; the d4-pawn often becomes a long-term asset.
1. e4 e5 2. Nf3 Nc6 3. Bc4 f5 4. Nc3 fxe4 5. Nxe4 d5
Developing with 4. Nc3 lets Black take on e4 and then play the thematic pawn fork 5... d5, hitting the c4-bishop and e4-knight simultaneously. White can untangle without losing material, but every reasonable continuation concedes Black free development, the bishop pair or a broad pawn center. This is one of the most comfortable versions of the gambit for Black.
If you play the Italian as White and someone uncorks 3... f5 against you, the first rule is: do not get greedy. The theory-approved answer is 4. d3, calmly defending e4 and declining the gambit. Black’s whole opening is built on open lines and fast development; a closed center with d3 denies both and leaves the long-term weakening of 3... f5 as a permanent liability. After 4. d3 Bc5, resist 5. Ng5 — the f7 fork that wins games in the regular Italian loses here, because 5... f4 and 6... Qh4 generate a mating attack worth far more than the h8-rook. Develop normally instead: Nc3, O-O, and meet ...f4 with timely play in the center. Taking the pawn with 4. exf5 is playable but demands precision — after 4... e4 your knight gets pushed back to g1 and Black regains the pawn with active pieces, so only go there if you know the concrete follow-ups. The central strike 4. d4 is a respectable alternative that fights for the initiative immediately. Above all, remember that 3... f5 weakened Black’s king: keep an eye on Qh5+ ideas and the a2–g8 diagonal, castle without delay, and avoid pawn grabs that leave your kingside undeveloped. White who develops soberly and keeps the center keeps the advantage; White who hunts material on move five often does not reach move twenty.
Decline with 4. d3, keep the e4 strong point, and treat ...f5 as a weakening rather than a threat. Develop with Nc3 and castle short, avoid the Ng5 pawn-and-rook grab, and look to open the center at a moment when Black’s king is the more exposed one. If you accept with exf5, be ready for ...e4 hitting the knight and return material rather than fall behind in development.
Play for the initiative from move three. In the accepted lines, push ...e4 to evict the knight, build the center with ...d5, and regain the f5-pawn with ...Qg5 or ...Bxf5 while White is still undeveloped. Against 4. d3, switch gears: ...Bc5, ...f4 grabbing kingside space, ...d6 and ...Bg4, then attack on the f-file and kingside. Never drift into a quiet position a pawn down — the gambit only works at full speed.
The gambit is named after Eugène Rousseau, a French-born player active in the United States in the mid-19th century, when romantic counter-gambits like this were a normal part of master practice. As opening theory matured, 3... f5 was judged too loosening to survive accurate play and largely disappeared from serious tournaments, surviving mainly as a footnote in Italian Game theory. Online blitz and rapid chess revived it: against the enormous population of club players who play the Italian on autopilot, an off-book move on move three has real practical value, and the trap-laden lines around 5. Ng5 made it a favorite of streamers and gambit repertoires. Engines confirm the old verdict — White is better with best play — but at club time controls the gap between "best play" and what actually appears on the board is exactly the space this opening lives in.
Let's play the Rousseau Gambit
Let's learn the Rousseau Gambit! I don't even know where to begin with this opening... it is SO aggressive. It is SO trappy. Best of all, it is SO fun. We begin with pawn to e5.
Let's continue with the Rousseau Gambit
We'll defend our pawn with knight to c6 next. So far this is all standard stuff. The fun begins on the next move.
Let's execute the Rousseau Gambit
The next move is just disgusting. I guarantee your opponent has never seen it before and I guarantee you'll win a LOT of games with it: pawn to f5. Yup, we're giving away a pawn AND exposing our king. Trust me and do it :)
White defended their pawn with the knight. Pretty normal move, right? Looks harmless enough. But actually… nope. That little knight just lost white the whole game. I'm serious. Pawn takes pawn on e4.
Now we just play pawn to d5, forking the knight and bishop, and white can resign.
We're winning a piece. We're winning the game.
OK so we gambited our pawn, now what? Well, now we can push our pawn to e4 and attack that helpless knight on f3.
We scared the knight away, now let's scare the bishop. Pawn to d5 wins us the whole center.
Boring people would just capture the f5 pawn here. But we're not boring, are we? Queen to g5, attacking the undefended g2 pawn.
That's totally fine - pawn takes bishop on c6.
Time to grab the f5 pawn. We could take it with the bishop, but the queen’s sitting a little uncomfortably where she is. Queen takes f5 cleans things up and puts her on a better square.
We have absolutely smoked white here - just look at their development! Or should I say, their lack-of-development.
We’ve already messed up Black’s pawn structure, so now we can go after the f5 pawn. Our queen’s not looking too comfy on g5—she’s just wide open to attacks. So let’s solve both problems at once with queen takes pawn on f5.
Well done! We're just crushing white here because of our strong central control. The computer gives us a +1.5 point advantage
Now we COULD play bishop takes pawn on f4, to even out the material. But I have a better idea: let's play queen to g5 and attack white's undefended g2 pawn.
It’s time to capture the f5 pawn. Both the bishop and queen could take it, but the queen is exposed on g5 and needs to move soon. Queen takes f5 is best, solving both problems at once.
Two pawns in the center? Check. More pieces developed? Check. Going to win this game? Check.
Woahhhh. White is trying to be SNEAKY. Now that their queen is pinning our pawn to our king, we can no longer take the knight. Well not so fast, white! We can just play queen to e7 and all of a sudden, the threat on the knight is renewed.
White ran away like the cowards they are. Let's play knight to f6 here, to stop white's queen from ever coming to h5, and also to set up a pawn-to-d5 in the future.
Aaaaaaaand game over. We have a deadly move on white that will probably make them quit chess. Knight to d4 attacking the queen AND the c2 pawn.
Dang... white moved their queen AND defended knight to c2. The funny part? We can STILL play knight to c2, completely sacrificing our knight.
Do you see the best move? The move that will make white quit the game? Hint: it's pawn takes pawn on d3, opening up a CHECK to the king and ATTACKING the queen. That means no-matter what white does, that queen is ours.
Pawn. Takes. Queen.
I think you got it from here :D
It is a practical weapon, not an objectively best one. With accurate play — especially 4. d3 — White keeps an edge. But at club level it scores well because Italian players are out of their preparation on move three and the natural-looking moves often lose by force.
Declining with 4. d3 is the theory-approved answer. White keeps the e4 pawn, refuses to open lines, and treats 3... f5 as a long-term weakening of Black’s king. Accepting with 4. exf5 is playable but requires knowing concrete lines after 4... e4.
Structurally yes — Black offers the f-pawn against an e-pawn center, exactly like a King’s Gambit with colors reversed. The catch is that Black is a tempo down on the real thing, which is why precise, aggressive play is mandatory rather than optional.
After 4. d3 Bc5 5. Ng5, the standard Italian fork trick backfires. Black plays 5... f4 and 6... Qh4, sacrificing the h8-rook to the wandering knight, and the attack with ...Nf6–g4 and a sacrifice on f2 crashes through — in the main trap line White gets checkmated while still materially ahead, the stray knight stuck on h8.
It suits improving players who want to practice attacking chess and punish autopilot opponents. Beginners should be comfortable with basic king safety first, since 3... f5 exposes Black’s own king — checks on the e8–h5 diagonal must always be taken seriously.
Reading about an opening is step one. The trainer at the top of this page drills all 24 lines against the moves real opponents play — the first lines are free.
Train the Rousseau Gambit now