Check-raise strategy: when to check-raise the flop, solved

Ask most players what they check-raise on the flop and the answer is short: sets, two pair, maybe a big draw. A check-raise announces a monster, so you save it for one.

The solver disagrees on almost every count. We pulled the big blind's flop strategy against a button c-bet on ten flops, and on 8♠ 8♦ 3♣ it check-raises 36% of its range against a small bet. Nearly half of those raises — 47% of the raising combinations — have no pair and no straight draw. On K♠ 9♠ 4♠, the same player in the same spot check-raises 7%.

So the honest answer to "how often should I check-raise?" is "it depends on the board and the bet size, by a factor of five or more" — and the hands that do it are not the ones most people pick. This article covers how often, with which hands, at what size, and the patterns that let you get it roughly right at the table without memorising a single grid.

What is a check-raise?

A check-raise is checking to an opponent, then raising after they bet. It is completely legal in every regular game; the only exceptions are some home games that ban it by house rule.

The classic version, and the one this article is about, is the big blind on the flop. The big blind defends against an open, checks the flop to the preflop raiser, the raiser makes a continuation bet, and the big blind raises. Because the big blind acts first on every street, the check-raise is its main way of taking the betting lead back.

The pot on the flop is 5.5bb. Against the small c-bet the big blind can raise to 3.45bb or 6.21bb; against the large one, to 11bb or 19.8bb. Full method is at the bottom.

How often should you check-raise? Ten flops

The table is sorted by the last column: how often the big blind check-raises once you account for how often the button actually uses each c-bet size on that board.

Three patterns stand out, and they matter more than any single row.

1. Small c-bets get check-raised far more than big ones. On every flop, the big blind raises a 25%-pot bet more often than an 80%-pot bet — on 8♠ 8♦ 3♣, 36% against 24%; on K♠ 9♠ 4♠, 7% against 1%. A small c-bet usually means the button is betting a wide range, much of it weak. Raising it is also cheap: a raise to 3.45bb costs about 2bb more than calling, into a pot of nearly 7bb. A large bet comes from a narrower, stronger range and costs far more to raise.

2. Paired and dry boards get raised the most, not the wet ones. This is the one that surprises people. The common advice is to check-raise more on wet, connected boards and less on dry ones. Against the same 25% c-bet, the big blind check-raises 27% on J♠ J♦ 4♣, 36% on 8♠ 8♦ 3♣ and 24% on K♠ 7♦ 2♣ — but 16% on Q♠ J♦ T♣ and 18% on 7♠ 6♦ 5♣. Those dry boards are where the button bets small with almost its whole range, and a range that wide is easy to attack. The exception is ace-high: on A♠ 7♦ 2♣ the button holds most of the aces, and the big blind raises 17%.

3. High boards that favour the raiser barely get raised at all. A♠ K♦ 5♦ and K♠ 9♠ 4♠ come in at 4% and 6% overall. The button has far more of the strongest hands on those boards — it opened every AK and most of the big pairs, which the big blind would have 3-bet. With few strong hands of its own, the big blind has little to raise for value, so it has room for very few bluffs either.

What goes into a check-raise range

Break the check-raising range into hand types, weighted by how many combinations actually take the line, and the "save it for a monster" picture falls apart. These are the eight rainbow flops; the two-tone and monotone boards are left out because the data is per hand class, so it can't separate flush draws from the rest.

The shape is the same everywhere: the range is polarised. One end is strong hands — two pair, sets, trips, straights. The other end is hands that are not good enough to call comfortably but have something going for them as a raise. Medium-strength hands mostly call.

What fills the bottom end depends on the board:

• On dry and paired boards, there are few straight draws to raise, so the solver raises hands with backdoor potential and overcards instead. On K♠ 7♦ 2♣, 65s check-raises 85% of the time and A4s 68%. On 8♠ 8♦ 3♣, 97s raises 94%. These are hands that would mostly have to fold to further aggression anyway; raising lets them win the pot now and keeps the big blind's range from being only strong hands.
• On connected boards, the draws are the bluffs. On Q♠ J♦ T♣, straight draws and gutshots make up 59% of the check-raises, and hands with nothing make up none.

Look at where the kings sit. The big blind's best top pairs mostly call: KQo check-raises 46% of the time, K9s 43%, K8s 35%. They are strong enough to call down against a wide range, and raising them folds out exactly the weaker hands they beat. The big blind's best hands and its weakest playable hands raise; the hands in between call.

How big to check-raise

The solver gets two raise sizes in this tree: small (2.5 times the bet) and large (4.5 times). Which one it uses depends almost entirely on the board.

• Dry and paired boards: mostly the small raise. Against a 25% bet, the big raise is only 27–51% of check-raises on J♠ J♦ 4♣, 8♠ 8♦ 3♣, K♠ 7♦ 2♣ and A♠ 7♦ 2♣. Not much changes on the turn, and the button's range is wide and weak, so a small raise does the job.
• Connected and draw-heavy boards: almost always the big raise. On Q♠ J♦ T♣, 9♠ 5♦ 2♣, 7♠ 6♦ 5♣ and A♠ K♦ 5♦, 87–96% of check-raises are the big size. The turn can change everything on these boards, so the big blind charges the maximum while its draws still have fold equity and its strong hands are still ahead.

The same rule covers both: size up when the next card can change who is ahead.

Hands that don't do what you'd expect

A few individual results are worth knowing because they cut against habit:

• Pocket fours on J♠ J♦ 4♣, a full house, check-raises 92% of the time. Slow-playing feels natural with a hand this strong, but the button is betting almost its whole range here and much of it will pay off a raise.
• AJo on J♠ J♦ 4♣, trips with the best kicker the big blind can have, raises 67% and calls the rest.
• 4-3 suited on 7♠ 6♦ 5♣, a made straight, check-raises only 50% of the time. It is not the nuts — 9-8 makes a bigger straight — and calling keeps the button's bluffs in.
• Pocket nines on Q♠ J♦ T♣ raise 96%. It is a pair with an open-ended straight draw: a hand that hates calling and loves fold equity.
• A9s on 9♠ 5♦ 2♣ — top pair, weak kicker — check-raises 100% of the time, while on A♠ 7♦ 2♣ the big blind's top-pair hands such as ATs and A9s raise only 23%. On a low board, top pair is vulnerable to a lot of turn cards; on an ace-high board, almost nothing scary can come.

A check-raise strategy you can use tonight

You don't need ten grids to get most of the value. Five rules cover it:

1. Attack small c-bets. When the button bets small on a dry or paired board, it is betting a wide, weak range. Check-raise more here than anywhere else — a quarter to a third of your range on boards like K-7-2 and 8-8-3.
2. Respect big c-bets. Against an 80%-pot bet, raise much less; on most boards, call or fold.
3. Polarise. Raise your strongest hands and your best bluffs. Call with medium-strength hands — top pair, good kicker usually belongs in the calling range.
4. Pick bluffs that have something. Straight draws on connected boards; on dry boards, suited connectors and suited aces with backdoor draws. Not pure air with no way to improve.
5. Size by board. Small raises on dry and paired boards, big raises on connected ones.

And the exploitative adjustment: these numbers assume an opponent who folds, calls and re-raises correctly. Many players fold too often to a check-raise, especially after a small c-bet that was mostly air, so against them you can add bluffs. If someone rarely c-bets small with weak hands, raise less. GTO vs exploitative play covers when to make these deviations and when to hold back.

Questions

What is a check-raise in poker? A check-raise is checking when it is your turn to act, then raising after an opponent bets. It is most common from the big blind on the flop, against the preflop raiser's continuation bet. It is legal in all standard games.

How often should you check-raise the flop? In this solve, the big blind check-raises a 25%-pot c-bet between 7% and 36% of the time depending on the flop, and an 80%-pot c-bet between 1% and 24% of the time. Dry and paired boards get raised most; high boards that favour the raiser, such as A-K-5 and monotone K-9-4, get raised least.

What hands should you check-raise? A polarised range: strong hands such as sets, two pair and straights, plus bluffs that have potential, such as straight draws on connected boards or suited connectors with backdoor draws on dry boards. Medium-strength hands such as top pair, good kicker usually call.

How big should a check-raise be? The solver mostly raises to 2.5 times the bet on dry and paired boards and to about 4.5 times on connected, draw-heavy boards, where the next card can change who is ahead.

Should you check-raise or slow-play a set? Usually raise. On K♠ 7♦ 2♣ the big blind check-raises 77 and 22 every time against a small c-bet, and on J♠ J♦ 4♣ it raises a full house 92% of the time. The button is betting a wide range, and much of it will call a raise.

Is check-raising bad etiquette? No. It is a standard play in every regular game. A few home games ban it by house rule, and that is the only time it is off limits.

Should you check-raise top pair? Sometimes, and more on low boards than high ones. On 9♠ 5♦ 2♣, A9s check-raises every time; on K♠ 7♦ 2♣, KQo raises 46% and calls the rest; on A♠ 7♦ 2♣, weak-kicker aces raise only about a quarter of the time.

Go deeper

This is the other side of our flop c-bet study, which covers the same ten flops from the button's seat. The c-bet charts have the button's full strategy on many more boards (K-7-2 rainbow, for example), and the big blind's preflop calling range that every grid here starts from is on the big blind vs button defense chart. For the theory behind folding and bluffing frequencies, see minimum defense frequency and value bets vs bluffs; for a step-by-step walkthrough, see the guide to check-raising the flop. All free, no account.

Method: all figures come from one solver tree — 6-max cash, 100bb effective, button opens 2.5bb, big blind calls — decoded from our solver database on 4 October 2026. The button can c-bet 1.38bb or 4.4bb; the big blind can fold, call or raise 2.5× or 4.5× the bet. Range-level percentages are the solver's own frequencies for the big blind's whole range at that decision. The overall column weights the two c-bet sizes by how often the button uses each. Per-hand figures are the solver's mix for each hand class, and the composition table weights each class by its combinations on that board and by how often the big blind calls the button's open with it preflop. Hand classes ignore suits, so flush draws are not separated; the two flush-draw boards are left out of that table. Percentages are rounded.