iGaming Journalist & Crypto Casino Analyst
Ask how many people have a gambling problem and you will get answers ranging from 0.4% to 15% of the adult population, depending entirely on which country, which screening tool and which threshold the number came from. This reference page collects the best available problem gambling prevalence data — global meta-analysis, national surveys and regulator statistics — and, just as importantly, explains why the figures differ so much and how to compare them without misleading your readers.
Key Findings
A 2024 Lancet Public Health meta-analysis covering 68 countries found 46.2% of adults gambled in the past year and 1.41% met criteria for problem gambling — roughly 80 million adults worldwide. National surveys diverge widely: Great Britain reports 2.4% at PGSI 8+, the United States estimates about 1% with severe problems, and Australia records 15% of adults experiencing gambling-related harm. Online casino and slot products carry the highest measured risk.
Why Prevalence Numbers Disagree So Much
Before any cross-country table is useful, three sources of variation have to be understood. Skipping them is how a 1.4% global figure and a 15% Australian figure end up in the same sentence as though they measured the same thing.
Different screening instruments
The three dominant tools are the Problem Gambling Severity Index (PGSI, nine items), the South Oaks Gambling Screen (SOGS, twenty items) and DSM-5 criteria for gambling disorder. They were built for different purposes and do not produce interchangeable rates. SOGS was designed in a clinical setting and tends to produce higher population estimates than PGSI; DSM-5 gambling disorder is a diagnostic category and is narrower than either.
Different thresholds within the same instrument
Even within PGSI, results depend on where the line is drawn. PGSI 8+ is conventionally read as "problem gambling"; PGSI 3+ captures moderate risk as well; PGSI 1+ captures anyone reporting any symptom. Sweden's headline rate uses PGSI 3+, Great Britain's uses 8+. Those are different questions with different answers, and neither is wrong.
Different denominators and recall windows
Some surveys report prevalence among all adults, others among people who gambled in the past year — a much smaller base that mechanically produces a higher percentage. Recall windows range from past four weeks to past twelve months. Survey mode matters too: online panels typically return higher gambling and problem-gambling rates than telephone or face-to-face interviews, which is why Great Britain's methodology change in 2023 produced a step-change that the regulator explicitly warns against comparing to older figures.
The Global Baseline
The most rigorous single source is the systematic review and meta-analysis published in The Lancet Public Health in 2024. It screened 3,692 reports and pooled 380 representative samples from 68 countries and territories, covering roughly 41% of the world's adult population and 50% of its adolescent population.
| Measure | Estimate | Population |
|---|---|---|
| Gambled in past 12 months | 46.2% | Adults |
| Gambled in past 12 months | 17.9% | Adolescents |
| Gambled in past 12 months — men | 49.1% | Adults |
| Gambled in past 12 months — women | 37.4% | Adults |
| Problem gambling | 1.41% | Adults |
| Any risk gambling — men | 11.9% | Adults |
| Any risk gambling — women | 5.5% | Adults |
| Estimated global burden | ≈ 80 million | Adults with gambling disorder or problematic gambling |
Two caveats travel with these numbers. The evidence base is heavily weighted toward high-income countries, with no estimates available from Pacific Island states, Central Asia or the Caribbean — so the global figure is really a high-income-country figure extrapolated. And "any risk gambling," which describes occasional experience of at least one behavioural symptom or adverse consequence, is far broader than clinical disorder; conflating the two is the most common misreading of this study.
Problem Gambling Rates by Country
The table below collects headline national figures. The instrument and threshold column is not decoration — it is the only way these rows can be read against each other honestly.
| Country | Headline rate | Instrument / threshold | Source and period |
|---|---|---|---|
| Global (68 countries) | 1.41% | Pooled, mixed instruments | Lancet Public Health meta-analysis, 2024 |
| Great Britain | 2.4% | PGSI 8+ | Gambling Survey for Great Britain, 2025 |
| United States | ≈ 1% severe; 2–3% mild/moderate | Composite national estimate | National Council on Problem Gambling |
| Australia | 15% experiencing gambling harm | Harm-based measure, not clinical | National gambling study, 2024 |
| Australia (clinical) | ≈ 1.4% moderate/severe | PGSI-based | National gambling study, 2024 |
| Sweden | 1.3% | PGSI 3+ | National survey, 2021 wave |
| Finland | ≈ 3.3% | Varies by wave | Comparative literature |
| Norway | ≈ 1.4% | Varies by wave | Comparative literature |
| Netherlands | ≈ 0.6% | Past-year problem gambling | Comparative literature |
| Germany | 0.4%–0.64% | Past-year problem gambling | Comparative literature |
Read down that column and the apparent 37-fold spread between Germany and Australia largely dissolves. Australia's 15% is a harm measure covering anyone who reported any gambling-related harm; its clinically comparable figure is about 1.4%, which sits almost exactly on the global pooled estimate. Great Britain's 2.4% is a PGSI 8+ figure from an online-push-to-web survey. Germany's 0.4% comes from an older telephone-based methodology. The instruments explain most of the variance; national culture and product availability explain the rest.
Great Britain: The Most Comprehensive Ongoing Survey
The Gambling Survey for Great Britain is one of the largest dedicated gambling surveys in the world, run by the National Centre for Social Research with the University of Glasgow on behalf of the Gambling Commission. Its 2025 annual report drew on 20,775 respondents and, for the first time, provided three years of comparable trend data.
- PGSI 8+ rate: 2.4% in 2025, down from 2.7% the previous year and 2.5% in the first wave — described by the regulator as broadly stable.
- Participation: 47% of adults gambled in the past four weeks; 27% excluding those who only played lottery draws.
- Sample size: around 20,000 respondents annually.
The Commission publishes explicit guidance warning that GSGB figures should not be compared with the earlier Health Survey for England series, because the change in survey mode alters the results. That guidance is routinely ignored in press coverage, which is how "problem gambling has tripled" headlines get written from a methodology change.
United States: A Fragmented Picture
The US has no equivalent of the GSGB. National estimates come primarily from the National Council on Problem Gambling, whose NGAGE survey series is the closest thing to a recurring national instrument.
- Approximately 2.5 million US adults (about 1%) are estimated to meet criteria for a severe gambling problem in a given year.
- A further 5–8 million adults (2–3%) are estimated to have mild or moderate gambling problems.
- Around 8% of American adults — close to 20 million people — reported experiencing at least one indicator of problematic gambling behaviour "many times" in the past year.
- NCPG estimates the annual national social cost of problem gambling at $14 billion, covering criminal justice and healthcare spending, job loss, bankruptcy and related consequences.
- Only a small minority of affected people — on the order of 8% — seek help.
The gap between the 1% severe figure and the 8% "any indicator" figure is where most US media confusion originates. They are both accurate and they describe very different populations. Note also that US prevalence research is contested: some analysts argue the underlying rate has been comparatively stable despite the rapid expansion of legal sports betting and online casino gaming since 2018, while others point to sharply rising helpline volume as evidence of growing harm. Both observations can be true if legalisation is shifting existing play into visible, help-seeking channels rather than only creating new players.
Australia: The Highest Losses Per Adult in the World
Australia is the standing outlier in gambling expenditure, largely because electronic gaming machines are widely available in pubs and clubs rather than confined to casinos.
- Australians lost an estimated A$32 billion (about US$21.1 billion) to gambling in 2024 — more than $1,200 per adult, the highest per-capita loss of any country.
- 65% of Australian adults gambled at least once in the past year.
- 15% experienced gambling-related harm, up four percentage points since 2019.
- Participation in high-risk gambling behaviours rose to 13.6% in 2024 from 11.6% in 2023.
The per-adult loss figure is the most-cited gambling statistic in the world and also the most frequently mangled. It is total losses divided by the total adult population, including the roughly one third of adults who do not gamble at all — so it is not the average loss of a gambler.
Product Risk: Not All Gambling Is Equal
The Lancet meta-analysis found that rates of problematic gambling were highest among people whose primary activity was online casino or slot play. This is the single most policy-relevant finding in the literature, and it maps onto a structural explanation: continuous products with short event frequency, immediate resolution and the ability to restake winnings within seconds produce a different risk profile than weekly lottery draws.
The practical consequence is that participation rates and harm rates do not track each other by country. A country where most gambling is lottery-based can have very high participation and low harm; a country with widespread machine or online slot play can show the reverse. Any ranking of countries by "gambling problem" that ignores product mix is close to meaningless. This is also why skill-based games sit awkwardly in the same datasets — the risk profile of a weekly strategic poker session differs structurally from continuous slot play, though both fall under the same regulatory umbrella.
Economic and Social Cost
Prevalence rates alone understate the policy stakes, because harm concentrates. The NCPG's $14 billion annual US social cost estimate is the most-cited figure, encompassing criminal justice costs, healthcare, lost productivity, job loss and bankruptcy. Comparable national cost studies exist for Australia, Great Britain and several Nordic countries, but they use different cost categories and are not directly comparable to one another. Researchers should treat cross-country cost comparisons with even more caution than prevalence comparisons.
The other consistent finding across jurisdictions is the help-seeking gap. Whatever the prevalence rate, the share of affected people who reach formal treatment is small — single-digit percentages in most studies. Any policy analysis that infers the size of the affected population from treatment volume will understate it by an order of magnitude.
Methodology
This article compiles published prevalence estimates; it does not conduct original survey research. Global figures are taken from the 2024 systematic review and meta-analysis published in The Lancet Public Health, which pooled 380 representative samples across 68 countries. National figures are taken from official regulator or government statistical publications where those exist — the Gambling Commission for Great Britain and the national gambling study for Australia — and from the National Council on Problem Gambling for the United States, which is the standard national reference in the absence of a federal survey.
Where a country's most recent figure comes from comparative academic literature rather than a current national survey, we have marked the instrument as varying by wave and treated the value as indicative. Rates measured on different instruments, thresholds, recall windows or survey modes are not statistically comparable, and we have deliberately kept the instrument column adjacent to every number rather than presenting a clean-looking ranked list that would invite false comparison. Figures were current as of August 2026.
Frequently Asked Questions
What percentage of the world's population has a gambling problem?
The 2024 Lancet Public Health meta-analysis estimated 1.41% of adults meet criteria for problem gambling, corresponding to roughly 80 million adults worldwide. Coverage is weighted toward high-income countries.
Which country has the highest problem gambling rate?
There is no defensible single answer, because national surveys use different instruments. Australia reports the highest gambling losses per adult in the world and 15% of adults experiencing gambling harm, but its clinically comparable moderate-to-severe rate of about 1.4% is close to the global average.
What is the problem gambling rate in the UK?
2.4% of adults scored 8 or above on the PGSI in the 2025 Gambling Survey for Great Britain, based on 20,775 respondents. The Gambling Commission describes the rate as stable across three survey years.
How many Americans have a gambling problem?
Around 2.5 million US adults are estimated to have a severe gambling problem, with a further 5 to 8 million experiencing mild or moderate problems. About 8% of adults report at least one indicator of problematic gambling behaviour many times a year.
Why do problem gambling statistics vary so widely?
Because they measure different things: different screening instruments (PGSI, SOGS, DSM-5), different severity thresholds, different recall windows, different denominators and different survey modes. Comparing headline percentages without checking those four variables is the most common error in gambling reporting.
Sources
- The prevalence of gambling and problematic gambling: a systematic review and meta-analysis — The Lancet Public Health (2024)
- The Lancet Public Health Commission on gambling
- Gambling Survey for Great Britain — Annual Report 2025, Gambling Commission
- National Council on Problem Gambling — FAQs and NGAGE survey findings
- Australian Institute of Family Studies — National gambling study
Cite This Article
If you use data from this article, please link back to https://www.deucescracked.com/blog/problem-gambling-prevalence-by-country-data. Suggested citation: DeucesCracked Editorial, "Problem Gambling Rates by Country: The Data," DeucesCracked, 24 August 2026. Please also credit the underlying primary sources listed above, particularly the Lancet Public Health meta-analysis and the Gambling Commission, when quoting specific figures.
Getting Support
This page is a statistical reference, not clinical guidance. If gambling is causing difficulty for you or someone close to you, confidential help is available: in the United States, the National Council on Problem Gambling operates the 1-800-GAMBLER helpline; in Great Britain, the National Gambling Helpline is available free on 0808 8020 133; in Australia, Gambling Help Online operates a national service. Most licensed operators also provide deposit limits, time-outs and self-exclusion tools. Our guides section and editorial standards page explain how DeucesCracked approaches responsible gambling coverage, and further industry research is collected in our analysis archive.
Join the Conversation
Be respectful. No spam. Strategy discussion welcome.