Cognitive behavioral therapy for gambling addiction: what it does and what the evidence shows

Cognitive behavioral therapy for gambling addiction is a short, structured course of talking therapy, usually 6 to 10 sessions, that works on two things at once: the beliefs that make gambling feel winnable, and the habits, cues and access to money that put you in front of a bet. It has more research support than any other treatment for gambling disorder, and the UK's national clinical guideline tells services to offer it first.

The evidence has a clear shape and a clear gap. In the weeks after treatment, CBT cuts gambling severity, frequency and money lost, with effects trials describe as medium to very large. Beyond that the picture thins: the few trials that followed people for 9 to 12 months found the advantage over control groups was no longer statistically significant. CBT is the best-supported first move, not a one-time cure, and it sits inside the wider guide to gambling addiction treatment.

The beliefs CBT targets

Gambling-related cognitive distortions are specific, repeatable errors about chance, not general negative thinking. A 2012 review in Psychology of Addictive Behaviors traced most of them to two ordinary mental shortcuts: judging by what a pattern resembles, and judging by what comes to mind easily. In a betting app, both misfire.

The near miss is the one most people underestimate. In a Cambridge study published in Neuron in 2009, near misses on a simplified slot machine were rated as less pleasant than plain losses yet increased the desire to keep playing, and they activated the same striatal and insula circuitry as real wins.

The behaviour CBT changes: triggers, money, urges

Belief work alone is not enough: the same 2012 review found that therapies aimed narrowly at correcting gambling beliefs had more limited success than cognitive restructuring delivered inside a broader CBT package. The behavioural side is built from components that recur across protocols:

• Trigger mapping. A written list of the situations, times, moods and cues that precede a bet: payday, the walk past a shop, a losing Sunday, an odds notification at 11pm.
• Stimulus control. Changing the environment so the trigger arrives less often: blocking software, self-exclusion, deleting apps, a different route home. The UK guideline lists blocking tools and stimulus control among the skills taught during treatment.
• Money control. Reducing access to gambling funds through card blocks, moving the balance, or giving someone else oversight. Access to money is a process CBT protocols target directly.
• Urge management. Urge surfing is the standard technique: rather than fighting or obeying the craving, you watch it rise, hold and fall, on the observation that urges are time-limited. If your worst hour is late at night, see stopping a gambling urge at night.
• Alternative activities. Scheduling something concrete into the hours gambling used to fill, because an emptied evening is itself a trigger.

Sessions come with homework: thought records, trigger logs, spending records. In trials, how much of it people completed tracked with whether they stopped gambling.

One exercise: the ABC thought record

Cognitive restructuring usually starts with a written log. A systematic review of 39 gambling studies found that studies reporting their technique commonly used a written aid such as an ABC log to challenge those thoughts. One row looks like this.

• A, activating event: 10:40pm, phone in hand, a push notification about boosted odds on a match.
• B, belief: "I've followed the team news all week, so this one is close to free money."
• C, consequence: Excitement, a £60 in-play bet, then two more to get it back.

Then the part that does the work. Name the distortion: research effort feeling like control over an outcome you do not control. Write the evidence against it, meaning what the last ten "researched" bets actually returned, from the statement rather than memory. Write a replacement sentence you can believe ("reading team news changes my confidence, not the odds") and pair it with one behavioural change, here turning off the notification.

What a course of CBT looks like

The UK's NICE guideline on gambling-related harms (NG248, January 2025) sets out the format. Services should offer group CBT first, to reduce gambling severity and frequency, starting as soon as possible after diagnosis. Individual CBT follows if you do not want a group, if no suitable group can be formed, or if a group is judged unsuitable for you.

More or fewer sessions are possible. Two requirements apply to both formats: the therapy follows an evidence-based protocol rather than general counselling, and it includes a relapse-prevention component covering triggers and what to do if a relapse happens. That is why a course ends with a written plan, and it is the subject of gambling recovery and relapse.

Online CBT with therapist support is a real option where services offer it. In Swedish routine addiction care, an internet-delivered programme of one assessment module and nine treatment modules, with weekly written therapist messages, matched face-to-face care on post-treatment outcomes in a 2024 cohort study in the Journal of Medical Internet Research.

What the evidence shows, and what it does not

Two syntheses carry the weight. The Cochrane review of psychological therapies for gambling (Cowlishaw and colleagues, 2012, still the most recent on the question) pooled 14 randomised trials with 1,245 participants, eleven comparing CBT with a control condition. At 0 to 3 months after treatment, benefits ranged from medium for financial loss (SMD −0.52) to very large for symptom severity (SMD −1.82). Only one trial, with 147 people, compared groups at 9 to 12 months, and its effects were smaller and not significant. The authors rated the evidence very low to low quality and wrote that the durability of the gain is unknown.

A 2023 umbrella review in Clinical Psychology Review pooled five meta-analyses covering 56 studies and 5,389 participants. At post-treatment, CBT beat minimal or no treatment on gambling disorder severity (Hedges' g = −0.91), frequency and intensity; at follow-up the difference from control was again not significant. The authors still issue a strong recommendation for CBT as an empirically supported treatment, while noting the effect is far from uniform: some people improve a great deal, some do not change, and some get worse.

Two limits matter when reading those numbers. Almost everyone in these trials volunteered, so the results describe people who had already decided to act. And the fade at follow-up is not proof that CBT stops working: control groups improve over time too, and very few trials measured the long run.

Free self-directed options

If you are waiting for an appointment, or not ready to see a clinician, the resources below are free and use the same techniques. All checked in September 2026.

One caveat. In a study of 231 people randomised to Gamblers Anonymous referral, GA referral plus a cognitive-behavioural workbook, or GA referral plus eight sessions of individual CBT, the therapist-led sessions beat the workbook on some outcomes. A workbook is a reasonable bridge to treatment, not a replacement for it.

Finding a therapist, and what it costs

Three questions filter most private therapists: do they use a CBT protocol written for gambling rather than a general addiction one, does the course include relapse prevention, and how many sessions is it.

CBT with medication or peer support

NICE places medication after therapy, not instead of it: consider naltrexone to reduce gambling severity if a completed course of psychological therapy has not achieved the agreed outcomes, or if relapses keep happening despite one. It should be started by or under the supervision of a specialist, was recorded as an off-label use in January 2025, and the guideline says to consider continuing therapy alongside it. Details are in medication for gambling addiction.

Peer support runs in parallel rather than in sequence. NICE recommends offering it as an integral part of treatment to anyone who wants it, and in the 231-person trial above every arm included a Gamblers Anonymous referral, with CBT adding benefit on top rather than replacing it. What the meetings are like is described on our page about Gamblers Anonymous.