Roulette Bets and Payouts: Every Bet, What It Pays and Its True Odds

Chance per spin is the number of pockets the bet covers divided by 37. True odds compare the losing pockets with the winning ones: a split covers 2 pockets and loses on 35, so its true odds are 35 to 2, or 17.5 to 1, against a payout of 17 to 1. The same half-unit shortfall, scaled to each bet, gives every row the same 2.70%.

Payouts are quoted “to 1”, so the winnings come on top of your stake. A winning £10 split returns £180: £170 in winnings and the £10 chip. A payout written “for 1”, such as “18 for 1”, counts the stake in the total; the money is the same.

The average cost is the stake multiplied by the house edge. It is what the bet loses per spin over a long run, not what any single spin costs: one spin either wins the full payout or loses the stake. A £10 straight-up bet and a £10 bet on red both cost 27p per spin on average, even though one wins £350 once in 37 spins and the other wins £10 almost every other spin.

The grid has 36 numbers in twelve rows of three, with zero at the top. A street covers one row, for example 13-14-15, and pays 11 to 1. A six line sits on the outer edge where two rows meet and covers both, paying 5 to 1. Splits can be placed between any two numbers that touch on the grid, across or down, and zero can be split with 1, 2 or 3.

Two bets include zero by name. The trio covers 0-1-2 or 0-2-3 and pays 11 to 1, like a street. The first four covers 0-1-2-3 and pays 8 to 1, like a corner. On a single-zero wheel both carry the usual 2.70%. On an American wheel the equivalent five-number bet, 0-00-1-2-3, pays 6 to 1 and costs 7.89%, the worst bet on either wheel, as our European vs American roulette page explains.

Tables often set separate maximums for straight-up bets and for outside bets, because a single number can cost the casino 35 times the stake in one spin. The sign at the table gives both limits.

Because they win so often, outside bets produce steadier sessions: smaller wins, more of them, and slower swings in your balance. That makes them the usual choice for a first visit to the table, but they are not cheaper than inside bets, with one exception. At a table with la partage or en prison, the even-money bets lose only half when zero hits, which halves their house edge to about 1.35%. Columns and dozens do not get that benefit.

Zero is the outside bettor’s only enemy. On a single-zero wheel it comes up once in 37 spins on average, and every outside bet on the layout loses unless a zero rule applies. The full explanation of those rules is on our main roulette rules page.

Imagine staking £1 on every number for one spin: £37 on the table. Exactly one number wins, paying £35 plus its £1 stake, so £36 comes back and £1 stays with the casino. That £1 in £37 is the 2.70% edge, and it does not matter how the £37 is spread across the layout; the result is the same for any mix of bets.

This is why no combination of roulette bets has a lower edge than the bets it is made of. Covering 30 numbers wins often, but each win pays only what the covered numbers pay, and the shortfall is still one pocket in 37. Our house edge guide uses the same method for every other casino game, and our method page shows how each figure on this site is worked out.

Take tiers du cylindre: six chips, one on each of six splits.^{1 (source)} If 27 comes up, the 27/30 split wins 17 to 1 and the other five chips lose, so the bet returns 18 chips for 6 staked. Voisins du zéro uses nine chips, two of them on the 0-2-3 trio and two on the 25-26-28-29 corner, so it pays more when one of those numbers hits. The wheel order and the chip placement for each call bet are listed on our roulette rules page.

Online European roulette games show the same payouts in their help screens, as the Gambling Commission’s technical standards require for the rules and the chance of winning.^{2 (source)} If a game changes a payout, for example by adding random multipliers to straight-up numbers, the help screen shows the new pay table and the return to player for that version.