Credit Card Casinos: Which Doors the Ban Closed and What the Open One Costs

Credit cards at a casino: the short version

• At a British-licensed cashier the answer has been fixed since 14 April 2020. A licensee may not be paid for gambling with a credit card, and the rule follows the card into a wallet, through a middleman and into a phone. Announced that January, it covers every gambling product, in a shop or on a screen, with lotteries bought in person the one carve-out.
• The offshore casinos out there sit outside it, and their terms say less than people assume. Visa and Mastercard appear on offshore payment lines in October 2026, which tells you the card networks are accepted and nothing at all about whether a credit card will clear. That is a cashier’s own decision, and nothing published anywhere states it.
• Price the borrowing before you price the bonus. At an assumed 22.9% a year, £450 of losses cleared at the minimum payment takes 169 months and £678.43 of interest. The same £450 paid off at £60 a month is gone in 9 months for £36.27. The repayment habit, not the size of the loss, decides what it ends up costing.
• A declined card is the system working. At a licensed site a refusal means the rule is doing its job; at your bank it means your own controls are. Neither is a problem to be solved by finding a cashier that says yes.
• Debit is untouched, and so is a phone bill, though for quite another reason. Credit cards are what the rule names, and a phone account is not one of those, although payment law holds carrier billing to £40 a payment and £240 a month, and the regulator’s own look back at the ban treated a phone account as a kind of borrowing.

The single rule worth keeping: when the only card left that might go through is a credit card, the sensible response is to stop for the night rather than to hunt for a cashier that says yes.

Every route a card can take, and where each one stops

The ban is two sentences long and the second does most of the work, because it reaches a credit-card payment that has gone through a money service business on its way. That one clause closes the detours. Here is every route someone might try, with the licensed answer and the offshore answer side by side, so the difference between the two is visible in one place.

Two details in that table are worth drawing out. A wallet provider can tell a credit card from a debit card by the opening digits of the number, the issuer identification number, which is what makes the separate-pots arrangement workable rather than theoretical. And the debit row is the only one where the answer is the same on both sides of the licence, which is a reminder that the thing the ban removed was never the card network. It was the borrowing.

What a gambling balance costs to clear, three sizes and four speeds

A credit card turns an evening’s losses into a balance that charges interest until it is gone, and that is the cost the ban keeps away from British cashiers. It can be priced, so here it is priced. Every figure below assumes an annual rate of 22.9%, which works out at 1.733% a month, a minimum payment of 2.5% of what you owe or £5 if that is more, and a card used for nothing else.

Read the £450 row across and the shape of the problem appears. At the minimum it takes just over fourteen years and costs half as much again as it would cost to borrow the money twice over; at £60 a month it is gone inside a year for less than a tenth of that. Measured against the losses themselves, minimum-payment interest comes to 58% on £180, 151% on £450 and 182% on £900, so the bigger the balance the worse the ratio gets, because more of each payment is swallowed before it touches the debt.

The £30 column carries the other lesson. On £180 it clears things in seven months; on £900 it takes 43 months and costs £381.31, because a fixed payment that comfortably outruns the interest on a small balance barely outruns it on a large one. There is a point below which a fixed payment never clears a balance at all, and the only safe way to find it is not to be near it.

Why the smallest payment takes the longest

The arithmetic behind those 169 months is simpler than it looks, and worth seeing once because it explains every long row in the table.

£450 × 1.733% = £7.80 of interest, so £457.80 is owed, and 2.5% of that is a minimum of £11.44

Of that £11.44, the interest takes £7.80 and only £3.64 comes off what you owe. That is a little under a third of the payment doing any useful work, with roughly 68p in every pound going to the lender. Because the minimum is a percentage of a shrinking balance, the payment shrinks too, which is why the finishing line keeps moving away: each month the debt falls by about 0.77% of itself, and 0.77% of a smaller number is a smaller number.

The £5 floor only starts to help near the end. Two and a half per cent of £200 is exactly £5, so below a £200 balance the floor takes over and the repayment finally becomes a fixed sum rather than a shrinking one. On the £450 example that point arrives years in, which is why the last stretch of a minimum-payment debt is slow in a different way from the first.

None of this is a trick played by anybody. It is what a minimum payment is for: it is the smallest sum that keeps an account in good standing, not a plan for clearing a balance. The practical response is to treat the minimum as a floor you never use and to pick a figure of your own, because the gap between £30 and £60 a month on a £450 balance is £39.43 of interest and nine months of your life.

Nine months of £45, and the sixteen months that follow

One balance is the easy case. The harder one, and the commoner, is a habit: a modest amount lost most months and left on the credit card while the next month’s losses join it. Picture £45 added at the end of each month for nine months with nothing repaid along the way, which keeps the arithmetic clear, and then £75 a month until the balance is gone.

Nine months: £405 lost, £434.24 owed, with £29.24 of interest already charged

Seven payments of £75 finish it, so the whole episode runs sixteen months from the first loss to the last payment. The interest comes to £56.65 and the total repaid to £461.65, which is 14% more than was lost. Put another way, the debt outlives the gambling by seven months, and the last three payments are for evenings you will struggle to remember.

Two things about that example are deliberately gentle. The monthly loss is small, smaller than many people’s weekly spend on this, and the repayment starts promptly and never falters. Stretch either assumption and the figures move quickly: the same habit carried at the minimum payment instead would put £405 of losses into the 57-month shape of the £180 row and then some, because each month’s addition restarts the slow part of the curve.

There is also a quieter effect that only shows on a statement. A bingo habit is paid for in £5 and £10 top-ups rather than one obvious deposit, so the month’s total hides itself among forty small lines and the monthly bill becomes the first honest summary of what an evening in a room costs; the £10 deposit bingo guide works through what a small top-up really buys in tickets and time. Checking a statement against what you thought you spent is an unglamorous habit and a useful one.

What £45 of borrowed money actually buys

Before deciding whether something is worth borrowing for, it helps to know what it buys, and that is a question we can answer by simulation rather than by assertion. Take the £45 from the habit above and play it at 50p a spin for up to 300 spins, which is a long evening at a modest stake.

Each row below is 100,000 simulated sessions, every session starting on £45 and betting a flat 50p until the balance could not cover another spin or 300 spins had gone by. The two models are ours: the House model slot, medium volatility with a 96.1% return, a win on 26.94% of spins, a feature round about once in 145 spins and a top prize of 2,500 times the stake; and the Wild model slot, high volatility with a 95.8% return, a win on 15.0% of spins, a feature round about once in 260 spins and a top prize of 12,000 times the stake.

After 90 spins every balance in both models was still alive. By 180 spins, 84.7% of the House sessions were still staking against 66.1% of the Wild ones, and by 300 spins it is 57.1% against 38.9%. The ends of the evening look quite different too. On the House model the median session finished on £11.39 and 26.0% ended above the £45 they started with; on the Wild model the median finished on 41p, essentially nothing, and 19.2% ended ahead. A high-volatility game does not take more on average, it takes it faster and in fewer, larger pieces.

300 × £0.50 × (1 − 0.961) = £5.85    300 × £0.50 × (1 − 0.958) = £6.30

Played all the way through, the average cost of that evening is £5.85 on the House model and £6.30 on the Wild one. Averages are not what happens to anybody in particular, though, and that is exactly why borrowing fits badly here. A lender charges interest on what you actually lost, not on the average of what you might have lost, and in nearly half of the Wild sessions what you actually lost was the whole £45 before the evening was over.

The figures published with the ban, and what happened after

When the rule was announced in January 2020 the regulator put numbers beside it, and they are worth seeing in proportion rather than as headlines.

800,000 ÷ 24 million = 3.3%, about one gambler in thirty

The honest reading is mixed and worth stating as such. The headline finding of the review is encouraging: the majority of those who had gambled on credit cards before the rule made no attempt to get around it. The problem-gambling figure that helped justify the ban rests on a sample small enough that a few answers either way would move it noticeably, and the switching figure rests on seventeen people. None of that makes the rule wrong, and the 800,000 estimate is large in absolute terms whatever share of gamblers it represents, but a page that quoted 22% without saying 150 would be misleading you.

One detail from that review has lasted. It counted a mobile phone account as one of the kinds of borrowing it asked about, which is why this page treats a carrier-billed deposit on a monthly contract as a mild form of credit rather than as a payment method like any other. The payment-law limits of £40 and £240 are the only ceiling on it, and they are not gambling rules at all; how carrier billing works in practice is set out in the pay by phone guide.

Moving to a debit card: what changes, and what does not

For most people the switch from a credit card to a debit card changes exactly one thing that matters, and leaves the rest of the cashier experience alone.

What changes is whose money is at stake. A debit card spends a balance you already hold; a credit card borrows at the card’s rate, and everything in the two sections above follows from that single difference. What does not change is the machinery. Card networks carry both of them, the same merchant category code is attached to the payment, the same bank gambling block watches for it, and the same cashier screens accept it.

Speed does not change either, and where figures exist they are about cards rather than about credit. Where a card issuer has signed up to Visa’s fast-funds service it has thirty minutes from approving a payment to release the money, and the clearest published account of that at scale is Visa’s 2022 case study of Paddy Power Betfair: Visa Direct handled 40.6% of withdrawals there, with £398m going out to over 360,000 customers in real time, against waits of as long as five days before it arrived. A bank transfer over Faster Payments is usually there almost immediately, with the occasional wait of as much as two hours. Those numbers belong to the rail, not to any casino, and nothing in the offshore terms publishes a payout time.

Two further details are worth knowing if you were using a credit card to put a layer between yourself and the cashier. Loading a wallet from a debit card is not free: Skrill charges 1.00% to add money by debit or credit card, and Neteller takes 2.5% on most upload methods, so a wallet layer costs something even before the casino’s own terms apply. And a credit card inside a wallet is still a credit card, which means the layer does not change what the money is. The fuller comparison of card routes is in the Visa guide and the Mastercard guide, and the phone-wallet versions in the Apple Pay guide.

The one protection the ban hands you is quiet and real: at a licensed cashier the question of whether to gamble on borrowed money has already been answered, so it is one fewer decision to make at eleven at night.

Brakes worth switching on before the credit card comes out

A ban at the cashier does not stop somebody wanting to borrow, so the brakes that matter sit with your bank and your card provider rather than with the casino. All of them work on the same principle: putting time between an impulse and a payment.

Online gambling paid for by Visa card without the card present has to carry merchant category code 7995, the code set aside for betting, lottery tickets, casino chips and wagers at racetracks, and a bank’s gambling block is a rule watching for that code. Monzo blocks card payments to any firm filing under it and lets you fix the cooldown yourself when you switch it on, anything between 2 days and a full year, which is the longest delay any of these offers. Starling turns away card spending at most gambling merchants, online or in person, and says itself that it may not catch every gambling-related payment; its block needs 48 hours to lift. NatWest, describing its own feature in 2020, declined gambling payments as they were attempted and built in the same 48-hour wait before a removal took effect.

Monzo reckons around 90% of banks offer something comparable, so the question for your own is not whether it has one but how long the delay runs. Its own count is above 700,000 customers who have used the block; in January 2022 it put activations above 400,000, of whom fewer than one in ten had ever turned it off permanently. If you hold a credit card, the useful call to make is to its provider rather than to your bank, because a credit card you cannot use for gambling takes the whole question off the table.

Two brakes cover ground no card block reaches. One registration with Gamstop shuts every online operator holding a British licence, a scheme they have all had to join since 31 March 2020, though it reaches none of the casinos out at that end. Device blocking from GamBan works on the sites themselves rather than on the payments, which covers some of what Gamstop cannot.

Offshore cashiers print Visa and Mastercard, and that is all they print

Look at the payment line on any offshore cashier at that end and you will find Visa and Mastercard on every one of them, Apple Pay on every one, and at least one coin on every one. What you will not find anywhere is a statement about credit cards, and that silence is the accurate position rather than an oversight. The British rule binds Commission licensees; these are licensed in Anjouan, in Costa Rica and, most commonly of all, in Curaçao; so whether a credit card clears at any of their cashiers is a question only that cashier can answer, and the answer can change without notice.

What does not change is what the licence was carrying. Self-exclusion through Gamstop covers every British-licensed site at once and none of these. The invitation to choose a deposit limit before a first payment, in force here since 31 October 2025, is absent. So is age and identity checked ahead of any play, which has been the rule here since 7 May 2019. The ten-times ceiling on bonus wagering, in place since 19 January 2026, is absent too, and these packages ask 35x to 40x. The £5 limit on a slot spin, £2 where the player is aged 18 to 24, does not reach them, and neither do the design rules that put at least two and a half seconds between spins and took autoplay away. A withdrawal that cannot be reversed back into play, settled here since 31 October 2021, does not apply. And the eight-week complaints clock with an independent adjudicator at the end of it does not apply either.

British law aims the offence at the business rather than at the customer: a company serving British players while holding no Commission licence is the one breaking it, and nobody prosecutes the person playing. What that person gives up is the entire list above, and a credit card adds a second loss on top of the first rather than replacing it.

The offshore brands marketed hardest to British players show the pattern without anyone having to name a particular casino. One Curaçao company, Santeda International B.V., stands behind Velobet, Donbet, Rolletto, Goldenbet and Mystake. Every one of the five takes cryptocurrency, not one shows up on the British register or joins Gamstop, nothing limits what any of them asks in wagering, and the Commission can pursue none of them for you. NineWin and Kinghills arrived recently from Curaçao-registered firms that operate families of sister domains, with no public record anyone can check, and they find British players through pages built on the words not on GamStop. An Anjouan licence held by RBGAMING N.V. covers Rainbet, a casino built around coins, Bitcoin among them. Roobet and Gamdom, two more Curaçao names, began in the trade in gaming skins and hold no British licence, so a player here sending coins to either has nobody regulating the payment, nobody to appeal to and nothing a bank can reverse. Naming them says nothing whatever about whether any of them takes credit cards. The section on casinos outside GamStop reads the whole corner licence by licence.

Why a licensed cashier asks where the money came from

The credit-card rule answered one question about borrowed money at the cashier door. Everything else about where a deposit comes from is handled by asking, and the published enforcement record explains why British cashiers ask as much as they do. Penalties land on the company holding the licence rather than on the name above the lobby, so the figures below belong to licence holders.

The clearest case about affordability is 888’s, settled at £9.4m in March 2022: no affordability check ran until deposits reached £40,000, and a monthly deposit limit of £1,300 was accepted from an NHS worker whose pay was £1,400 a month. That limit is 93% of the month’s pay, which is the same question a credit card lets somebody avoid entirely. The William Hill group’s £19.2m in March 2023 turned on speed as much as size: a newly opened account spent £23,000 in twenty minutes unchecked, one customer lost £14,902 in seventy minutes, a £19,000 single bet drew no request for evidence, and 331 customers who had excluded themselves at Mr Green were not shut out across a sister licence. The regulator’s own description of the failings was widespread and alarming. Entain’s Ladbrokes Coral group paid £17m in August 2022 after one customer deposited £742,000 across fourteen months with no proper check on the source of it and another put in £524,501 in under a year. Kindred’s £7.1m in March 2023 covered 32Red and the company running Unibet here, and included a player who put in £43,000 and was £36,000 down within seven days, while most people under a source-of-funds request carried on depositing through the fortnight they had to answer it. Betway’s £11.6m in March 2020 concerned seven high-spending customers, one of whom deposited over £8m and lost over £4m across four years; £5.8m of that money was found to be, or suspected of being, the proceeds of crime. And the two licence holders behind bet365 were penalised £582,120 in April 2024 because a new customer could pay in before any financial sanctions screening had been done.

In June 2026 the Commission’s Director of Enforcement put the pattern into two complaints: warning signs overlooked in the bank statements and payslips customers had themselves handed over, and too much weight placed on fixed financial thresholds. So when a cashier asks how you fund your play, that is the reason, and it is precisely the question a credit card is designed not to raise.

Which cards a British brand takes is for its own cashier to display, and credit cards are not among them under any licence. Beyond that, brands differ in ways worth knowing. Admiral Casino, Genting Casino, Grosvenor Casinos and Betfred all keep premises in British town centres, which makes them the easiest of all to look over before you link a card; Betfred’s largest Commission settlement is £3.25m, in 2023. bet365 leads with its sportsbook and puts the casino second, carries Heart Bingo on the same licence, and makes a deposit the condition of a casino welcome as a matter of course. Flutter’s casino-first name is Sky Vegas, sitting beside the betting-led trio of Sky Bet, Betfair and Paddy Power. Gamesys runs one platform under Double Bubble Bingo, Rainbow Riches Casino, Monopoly Casino, Jackpotjoy and Virgin Games, where a welcome usually takes the deposit-ten-and-play-with-more shape. LeoVegas is mobile-first and carries Bet UK, Pink Casino and BetMGM, which opened here in 2023 using LeoVegas’s licence and platform. MrQ and PlayOJO promise no wagering on any bonus at all, which turns a win there into cash rather than a balance to turn over, and whichever you pick, the deposit behind it had to be made without a credit card. Casumo, Videoslots, NetBet, Mr Vegas, Midnite and Kwiff put the name of their licence holder, and the Commission account number beside it, at the foot of the site, which is where any check should begin.

Price your own balance and your own hour

The first calculator prices the play rather than the borrowing, and it opens on the evening simulated above: 300 spins at 50p with a 96% return. It gives the total staked, the average cost and the average returned, and the useful move is to change the stake rather than the number of spins, because the stake is the dial that actually shifts the bill.

Put the answer beside the repayment table and the comparison does the arguing for you. An evening of that size costs around £6 on average; carrying £450 of such evenings at the minimum payment costs £678.43 in interest. The cost of the gambling and the cost of the borrowing are two separate bills, and only one of them appears while you are playing.

The second calculator prices a welcome package, which is the other thing people reach for a card to fund. It opens on a £60 deposit with a 100% match and 35x on the bonus at a 96% return, the lightest wagering found at this end of the market, and shows the £2,100 of play-through those terms require. Change the wagering to 40 for the commoner setting and watch the required stakes climb while the bonus stays exactly where it was.

Running both in that order gives you three numbers to sit with: what an evening costs, what the offer asks for, and what the borrowing would add. If the third number is the one making the first two possible, the answer to the whole page is no.

Where these figures come from

The rule itself, how far it reaches, its wording about money service businesses, the e-wallet guidance behind the separate-pots arrangement and the figures published in January 2020, UK Finance’s estimate of 800,000 among them, all come from what the Gambling Commission published with its decision and its guidance. The review figures come from the Commission’s look back at the ban, dated 2 November 2021, which rested on a small survey. The crypto position comes from the Commission’s guidance on digital and virtual currencies, in its version of 5 July 2021, together with what it said publicly during 2026. The other dated rules, the enforcement outcomes and the June 2026 remarks by its Director of Enforcement are the Commission’s own. The merchant category code, the fast-funds window and the 2022 payout case study come from Visa, the Faster Payments timing from Pay.UK, the block descriptions from Monzo, Starling and NatWest, the wallet loading charges from Skrill’s and Neteller’s fee pages, the phone-billing caps from payment law, and every brand fact from this site’s brand sheet.

Every repayment figure here comes from a month-by-month model with stated assumptions: an annual rate of 22.9%, converted to a monthly rate of 1.733% as the twelfth root of 1.229 minus one; a minimum payment of 2.5% of the balance or £5, whichever is larger; interest charged before each payment; and no fees, no further spending and no change of rate. The habit example adds £45 at the end of each of nine months with nothing repaid, then repays £75 a month until the balance is clear. The rate is an illustration rather than a quotation, and a card charging more or less will move every figure in proportion.

The simulated figures are ours and describe models, not real games. Each row is 100,000 simulated sessions starting from £45, betting a flat 50p a spin for up to 300 spins or until the balance could not cover another, on our House model slot (medium volatility, a 96.1% return, a win on 26.94% of spins, a feature round about once in 145 spins, a top prize of 2,500 times the stake) and our Wild model slot (high volatility, a 95.8% return, a win on 15.0% of spins, a feature round about once in 260 spins, a top prize of 12,000 times the stake). An average cost here is spins times stake times the house edge, which is one minus the return.

Questions about cards, bans and borrowing

Next on your sheet

• Ways to depositSixteen ways to pay in, priced and capped side by side
• VisaDebit against credit, code 7995 and what fast funds promise
• MastercardCard deposits and the routes winnings take back
• Pay by phoneCarrier billing, the two caps and the nearest substitutes
• PayPalWhere the button really works, and what stands in for it
• Outside GamStopWhat an offshore licence leaves out, read before you deposit