Responsible Gambling Statistics 2026: Harm & Protection Data

2.4% of adults in Great Britain scored eight or more on the Problem Gambling Severity Index in 2025, down from 2.7% in 2024, with a further 3.5% scoring three to seven, according to the third annual Gambling Survey for Great Britain. Depending on the population base applied, that headline rate corresponds to roughly 1.3 to 1.4 million adults, and different organisations publish figures at both ends of that range. Over the same period GAMSTOP, the British online self-exclusion scheme, passed 562,000 total registrations, and the US National Problem Gambling Helpline handled more than 31,000 contacts a month. This page compiles the responsible gambling data that journalists, researchers and compliance teams cite most often: prevalence measures and their methodological caveats, self-exclusion scheme numbers, helpline demand, funding for research, prevention and treatment, the contested evidence on advertising restrictions, and the obligations that responsible gambling rules place on operators and their affiliate partners. It is reviewed quarterly.

Problem Gambling Prevalence in Great Britain: 2.4% at PGSI 8 or More

2.4% of adults in Great Britain scored eight or more on the Problem Gambling Severity Index in 2025, a decline from 2.7% in 2024, with a further 3.5% scoring in the three to seven moderate risk band, per the third annual Gambling Survey for Great Britain. The survey is commissioned by the Gambling Commission and carried out by the National Centre for Social Research with the University of Glasgow, using a push-to-web methodology across four quarterly waves and a base of 20,775 adults, and 2025 was the first year providing three comparable annual data points. The population equivalent of the headline rate is published differently by different organisations: reporting on the survey has cited approximately 1.3 million adults, while the charity Gambling With Lives has published 1.4 million. Both figures derive from the same 2.4% rate applied to slightly different adult population bases, and this page shows the range rather than selecting one.

One caveat should accompany every citation of these figures. The Gambling Commission has repeatedly advised that GSGB results are not directly comparable with the earlier prevalence estimates produced through health surveys, because the survey mode and question sequence changed, and that the new series should be treated as a new baseline rather than a continuation of the old one. Year-on-year movements within the GSGB series, such as the fall from 2.7% to 2.4%, are the appropriate comparison. Comparisons between GSGB figures and pre-2023 health survey figures are not, and a decline reported within the GSGB series should not be characterised as a decline from the historic health survey rates.

Prevalence Beyond Britain: 3 Reasons Comparisons Mislead

Three methodological differences make cross-country problem gambling comparisons unreliable: the screening instrument, the survey mode, and the reference period. Great Britain uses the PGSI in an annual push-to-web survey with a large national base, which makes it the most continuous national series in the world. The United States has no equivalent annual national prevalence survey, and US estimates in circulation derive from state-level studies conducted in different years with different instruments, so a single national US rate should be treated with caution wherever it appears. Across the European Union, prevalence studies are conducted by member states on their own schedules using several different screens. This page therefore does not present a country-by-country prevalence league table, because assembling one would require comparing figures that are not comparable.

• Screening instruments differ: the PGSI, the DSM-based criteria and the SOGS produce different rates from the same population, so an instrument change alone can move a headline figure
• Survey mode matters: push-to-web, telephone and in-person interviewing produce different response profiles, which is the core reason the Gambling Commission warned against comparing GSGB with earlier health survey figures
• Reference periods differ: some studies ask about the past 12 months and others about lifetime experience, which are not the same measure
• The United States has no annual national prevalence survey comparable to the GSGB, so continuous US indicators come from helpline and treatment demand rather than from population prevalence
• Where a single global problem gambling rate is quoted, it is almost always an average of incomparable national studies rather than a measurement, and this page does not publish one

Self-Exclusion: 562,000 GAMSTOP Registrations by the End of 2025

562,000 people had registered with GAMSTOP by the end of 2025, after 117,756 new registrations during the year, a 19% year-on-year increase and the scheme's highest annual total. Monthly records were set twice in succession: 10,281 registrations in April 2025, the first month ever above 10,000, then 10,344 in May. Growth was driven disproportionately by younger consumers, with registrations among 16 to 24 year olds rising 44% year on year in the first half of 2025. The composition of exclusion lengths also shifted, with six-month exclusions up 29% year on year, although the five-year maximum remains the most commonly chosen option overall at about 47% of all exclusions, and is the most popular choice in every age group except the under-25s.

Rising self-exclusion numbers are not straightforwardly a measure of rising harm. They reflect some combination of harm prevalence, scheme awareness, referral practice by operators and treatment services, and the growing number of consumers who use self-exclusion as a precautionary control rather than after a crisis. GAMSTOP's own reporting on the shift toward shorter six-month exclusions is consistent with more precautionary use. What the numbers do establish unambiguously is scale: more than half a million people in Great Britain have chosen to block their own access to licensed online gambling, and any operator or marketing partner working in that market is operating in a population where that is true.

Helpline Demand: More Than 31,000 US Contacts a Month

The US National Problem Gambling Helpline handled more than 31,000 contacts per month in 2025, and 49.48% of them came from people aged 18 to 34, per the National Council on Problem Gambling's annual helpline report. The profile of those contacts shifted in ways that track the growth of online gambling. Financial difficulty was the most frequently cited reason for making contact, reported by more than 73% of contacts, up from 66% in 2024. Contacts citing online and app-based gambling rose to 31% from 23% a year earlier, approaching the level reported for casino gambling. The helpline has also been rebranding its access route, with 1-800-MY-RESET introduced alongside the existing 1-800-522-4700 and 1-800-GAMBLER numbers, which remain active.

Helpline data is a demand indicator, not a prevalence measure. It counts people who sought help, which is influenced by awareness campaigns, referral pathways and the visibility of the helpline number in operator and broadcast advertising, as well as by the underlying level of harm. It cannot be converted into a national prevalence rate. Its value is that it is continuous, it is collected the same way each year, and in the absence of an annual US national prevalence survey it is the closest thing the United States has to a year-on-year harm indicator.

Funding for Research, Prevention and Treatment: the 1.1% Statutory Levy

Britain's statutory gambling levy commenced on 6 April 2025 at a rate of 1.1% of gross gambling yield for online operators, with rates varying by licensed activity, and is estimated to raise GBP 90 to 100 million in 2025/26. The levy replaced the previous voluntary contribution arrangement and is allocated in fixed proportions: 50% to NHS England and the appropriate bodies in Scotland and Wales to commission treatment and support, 30% to prevention with a cap of up to GBP 30 million a year, and 20% to UK Research and Innovation for a dedicated gambling research programme. The structural significance is that funding for treatment and research is no longer set by operator discretion, and that the research strand sits with an independent public funder rather than with industry-nominated bodies.

Advertising Restrictions: 2 Bodies of Evidence That Disagree

Two bodies of evidence on Britain's whistle-to-whistle television advertising ban point in different directions, and any honest summary has to present both. Analysis cited by the Betting and Gaming Council, conducted by Enders Analysis, found a 97% reduction in the number of television betting adverts seen by children during the ban's first 12 months of operation. Academic work published since, including 2025 research from the University of Bristol, has found that football audiences were exposed to more than 27,000 gambling messages across the opening weekend of the Premier League season, close to triple the level recorded in 2023, because pitch-side advertising, shirt sponsorship and in-broadcast branding fall outside the ban's scope. Both findings can be true simultaneously: the ban addressed a specific advertising format and did not address total exposure.

• The whistle-to-whistle ban restricts television betting advertising around live sport broadcasts and does not cover pitch-side hoardings, shirt sponsorship or in-programme branding
• Enders Analysis found a 97% reduction in TV betting adverts seen by children in the ban's first 12 months, as cited by the Betting and Gaming Council
• University of Bristol research in 2025 recorded more than 27,000 gambling messages during a single Premier League opening weekend, close to triple the 2023 level
• Researchers have also noted that younger audiences frequently do not recognise sponsorship and branding as advertising at all, which complicates measurement of effective exposure
• No controlled evaluation linking the advertising ban to a change in measured gambling harm has been published, and this page does not claim one exists

Responsible Gambling Obligations for Affiliate Programs

Four obligations follow directly from the fact that operators remain legally responsible for the conduct of their affiliates: audience targeting controls, creative and disclosure standards, exclusion of self-excluded and vulnerable users, and documented due diligence on every marketing partner. Under UKGC licence conditions and codes of practice, a licensee is accountable for marketing carried out on its behalf, which means an affiliate breach is an operator breach. The MGA sets comparable licensee obligations, Germany's GGL enforces its own advertising restrictions, and in the United States the FTC endorsement guides govern disclosure while the ASA's influencer marketing rules apply in Britain. None of these frameworks treat affiliate marketing as a lesser category of advertising, and none of them accept lack of visibility into partner activity as a defence.

Commission design is part of compliance, not separate from it. Commercial models built on NGR, which is GGR after permitted deductions, or on RevShare, CPA or hybrid terms, all reward acquisition and player lifetime value, and regulators have made clear that the existence of that incentive does not reduce an operator's duty to protect consumers. Practical consequences follow. Qualification rules should exclude any player who has self-excluded or been flagged for intervention, and commission should not accrue on their activity. Geo-targeting must be enforced at the link level so that traffic cannot reach markets where the operator is unlicensed or where a specific product is restricted. Negative carryover and other clauses that increase pressure on partners to chase volume should be reviewed for the behaviour they encourage. And the same fraud controls that detect bonus abuse, multi-account signups and self-referral schemes also surface patterns worth reviewing from a player protection perspective, because unusual deposit and account behaviour is relevant to both questions. Track360's operator guidance on responsible gambling marketing and affordability tooling covers the implementation detail.

Methodology & Sources

Two source classes cover every figure on this page: named regulator, government, scheme and research publications, and clearly flagged ranges where published figures differ. This page contains no Track360 estimates. It follows a fixed process.

1. Prevalence figures come from the Gambling Survey for Great Britain, commissioned by the Gambling Commission and conducted by NatCen with the University of Glasgow. We reproduce the Commission's own caution that GSGB results are not directly comparable with earlier health-survey-based estimates, and we do not present them as a continuation of the earlier series.
2. Where organisations publish different population equivalents for the same rate, as with the 1.3 to 1.4 million adult figures derived from the 2.4% PGSI rate, we present the range and name both sources rather than selecting one.
3. Self-exclusion figures come from GAMSTOP Group annual reporting and its published updates. These cover Great Britain's online self-exclusion scheme only and are not comparable with land-based or national schemes elsewhere.
4. US helpline figures come from the National Council on Problem Gambling's annual helpline reporting. Helpline contacts measure help-seeking, are influenced by awareness and referral practice, and cannot be converted into a prevalence rate.
5. Levy figures come from the UK Government consultation response and Gambling Commission guidance on the Gambling Levy Regulations 2025. The GBP 90 to 100 million figure is a government estimate for 2025/26, not an outturn.
6. Advertising evidence is presented as two conflicting bodies of work, the Enders Analysis finding cited by the Betting and Gaming Council and subsequent academic research including 2025 University of Bristol work, because they measure different things and reach different conclusions. We do not claim that a controlled evaluation linking the ban to measured harm outcomes exists.
7. We publish no global problem gambling prevalence rate and no country-by-country prevalence table, because national studies use different instruments, survey modes and reference periods and are not comparable.
8. This page is reviewed quarterly, within 30 days of GSGB, GAMSTOP and NCPG publications; the next scheduled review follows the Q3 2026 releases.

Frequently Asked Questions

Five questions cover the responsible gambling data points journalists, researchers and compliance teams request most often.