Gambling addiction statistics: prevalence by country and what each number measures

About 1.41% of adults worldwide gamble at a problematic level, and a further 8.7% gamble at some level of risk. Those are pooled estimates from 366 representative samples in 68 countries, published in The Lancet Public Health in 2024. National surveys land either side of that figure: 2.4% of adults in Great Britain scored 8 or more on the PGSI in 2025, and 2.2% of adults in Germany met DSM-5 criteria for a gambling disorder in the same year.

Those three numbers come from three different measuring sticks. A PGSI score of 8 or more, four or more DSM-5 criteria, and “at least one problematic behaviour many times in the past year” are not interchangeable, and much of the spread between countries is measurement rather than behaviour. Every figure on this page therefore carries its year, its jurisdiction, the screen it came from and a link to the source.

How many people have a gambling addiction worldwide

The broadest global estimate is a systematic review and meta-analysis by Tran and colleagues, published in The Lancet Public Health in July 2024. It screened 3,692 reports and pooled 366 representative population samples from 68 countries and territories, covering studies published between January 2010 and March 2024.

• 46.2% of adults (95% CI 41.7–50.8) and 17.9% of adolescents (14.8–21.2) had gambled in the past 12 months.
• 8.7% of adults (6.6–11.3) were classified as engaging in any-risk gambling.
• 1.41% of adults (1.06–1.84) were classified as engaging in problematic gambling.
• 15.8% of adults who gambled on online casino games or slots (10.7–21.6) were classified as problematic — the highest rate for any single activity in the review.

The World Health Organization's gambling fact sheet, published on 2 December 2024, puts the global figure a little lower, at 1.2% of the world's adult population with a gambling disorder, and adds a sex split drawn from the Lancet Public Health Commission: roughly 11.9% of men and 5.5% of women worldwide experience some level of harm from gambling. The two global numbers differ because a gambling disorder is a clinical diagnosis with a fixed criteria count, while “problematic gambling” is whatever a survey screen flags at its top threshold. The clinical definition is set out in what gambling addiction is.

Problem gambling rates by country

Ten national surveys, each with the screen it used and the year its data were collected. The rows are not a league table: Germany's 2.2% and Sweden's 0.5% are four years and one instrument apart, not a statement about German and Swedish behaviour.

Three cautions before quoting a row. Canada's percentages are shares of past-year gamblers, because that is how Statistics Canada published them; 64.5% of Canadians aged 15 and over gambled in 2018, so the population share is lower again. The US row is not a prevalence estimate at all: NCPG states that NGAGE “is not designed to assess the prevalence of gambling disorders, as defined by” the DSM-5, and separately estimates 2.5 million American adults likely to have a gambling disorder, with 5 to 8 million more showing some problematic behaviour.

Who gambles at risk: sex, age and format

Sex is the most consistent finding in the data. Globally, 49.1% of men and 37.4% of women gambled in the past year (Tran et al., 2024). The gap widens at the harmful end: 3.2% of German men versus 1.1% of German women met DSM-5 criteria in 2025; 0.8% of Swedish men versus 0.2% of Swedish women scored PGSI 8+ in 2021; and 10% of US men versus roughly half that share of US women reported at least one problematic behaviour many times in 2024.

Age runs the other way from most health risks. NGAGE 3.0 found 15% of Americans aged 18–34 met at least one problematic-play criterion in 2024, against 2% of those aged 55 and over. Finland's 2023 survey found the highest rate among 18- to 29-year-olds, at 6.9%, followed by 5.4% among 30- to 44-year-olds.

Format matters more than either. The Lancet Public Health meta-analysis found problematic gambling among 15.8% of people who gambled on online casino games or slots — more than eleven times the 1.41% adult rate. NGAGE 3.0 found the same pattern in weaker form: 19% of Americans who made at least some bets online met a problematic-play criterion, and 28% of those who bet on ten or more different activities did. NCPG's reading is that frequency and breadth of gambling predict problems more strongly than any single product.

Sports betting statistics since the 2018 US ruling

NCPG designed the first NGAGE survey in 2018 as a deliberate baseline, taken immediately after the Supreme Court ruling that let states legalise sports betting. Three waves now exist, which makes this the cleanest before-and-after series available for a gambling expansion.

Legal sports betting reached 38 states and Washington DC by September 2024, with online betting in 30 of them, yet participation did not keep climbing. NCPG's interpretation is that the 2021 spike in problematic play was probably driven more by the pandemic than by the expansion of sports betting, and that most adults who want to bet on sports do so regardless of their state's law. What did change is the bet type: parlays nearly doubled as a share of sports bettors between 2018 and 2024. Within the 2024 sample, 17% of traditional sports bettors and 24% of fantasy sports players met at least one problematic-play criterion. The behavioural pattern behind those numbers is described in sports betting addiction.

Harms recorded alongside gambling problems

Suicide is the harm with the strongest register-based evidence. A Swedish nationwide study of 2,099 people diagnosed with gambling disorder in specialist care between 2005 and 2016 found a 15-fold increase in suicide mortality and a 1.8-fold increase in all-cause mortality against the general population aged 20–74 (Karlsson & Håkansson, 2018). A population-based study of coronial records in Victoria, Australia found that 201 of 4,788 suicides between 2009 and 2016 — 4.2% — were gambling-related, 83% of them male; the authors note that coroners do not routinely investigate gambling, so this is an underestimate. These are associations in clinical and coronial populations, not a personal risk figure for any individual.

Financial and relationship harms show up in survey data. Among British adults who gambled in the 12 months to 2025, 6.4% had at least occasionally cut spending on everyday items because of their gambling, 5.8% had used savings or borrowed money, and 5.9% had lied to family about it; 2.7% reported at least one severe consequence, most often relationship breakdown (1.7%). Of all GSGB 2025 participants, 11.3% reported suicidal thoughts or attempts in the past year, and 4.6% of that group said it was related to their gambling a little or a lot.

Harm does not stop at the person gambling. The WHO estimates that for every person gambling at high-risk levels, an average of six other people, usually non-gamblers, are affected. In Finland's 2023 survey, 20% of respondents were concerned significant others of someone with a gambling problem. In Great Britain, 43.2% of GSGB 2025 participants said someone close to them gambles; within that group, 10.8% reported health problems including stress or anxiety and 3.3% relationship breakdown. If you are in that position, how to help someone with a gambling addiction covers what works.

How many people with a gambling problem get help

A meta-analysis of 24 representative studies published in Addiction in 2022 found that 20.63% of people with problem gambling (95% CI 12.89–31.35) had sought help of any kind — professional services, self-help or family and friends. Among people at moderate risk the figure was 3.73%, and among low-risk gamblers 0.27%. Across whole populations, current help-seeking was 0.14%, the figure the WHO fact sheet cites.

Put the way the authors put it: one in 25 moderate-risk gamblers and one in five people with problem gambling have sought help. Awareness is not the whole explanation: 36% of US adults had heard of the national helpline number in 2024, rising to 59% among people betting on seven or more activities. Americans reporting problematic behaviours were more than twice as likely as other gamblers to agree that people with a gambling problem are unlikely to recover (37% versus 15%), which is the opposite of what the treatment evidence shows; the options are set out in gambling addiction treatment.

Why the numbers disagree

Four mechanical differences explain most of the variation in the table above, and all four are documented by the agencies that publish the figures.

• The screen. PGSI 8+ and DSM-5 gambling disorder do not select the same people. Germany's DSM-5 rate counts mild cases at four criteria; Sweden's PGSI 8+ band has no equivalent tier. The US series uses neither, and NCPG says so in the report.
• Survey mode. The ESRI's Irish estimate of 3.3% is roughly ten times a 2019 measure, which the ESRI attributes to anonymous online surveying replacing face-to-face interviews rather than to a tenfold rise in problem gambling. Finland's THL ran the same comparison inside its 2023 study and concluded that mode differences of about 1.5 percentage points or less cannot be distinguished statistically.
• The base. Some agencies report shares of all adults, others shares of past-year gamblers. Canada's 0.3% is of gamblers; Australia's 1.23% is of all adults, and the same data gives 2.16% among gamblers.
• Age range and funder. Great Britain surveys 18 and over, Germany screens 18–70, Finland 15–74, Sweden 16–84. Funding differs too: the German survey is financed by the Deutscher Lotto- und Totoblock, the British one by the regulator.

The Gambling Commission attaches an explicit warning to its own data, and it applies to any comparison drawn across this page: “The push-to-web methodology of this survey means that estimates presented in this report are not directly comparable with results from prior gambling or health surveys and such comparisons should not be used to assess trends over time.” Within a single series, comparison is fair — the share of British adults scoring zero on the PGSI was 76.1% in 2023 and 76.9% in 2025, and the 8+ share held at 2.4%. Across series, it is not.

One limitation sits under all of it: these are self-reports about a behaviour people conceal. NGAGE 3.0 found a majority of US adults still attribute gambling problems to a lack of willpower or moral weakness, and 5.9% of British gamblers admitted lying to family about their gambling in 2025. Under-reporting is the expected direction of error.

Country-by-country helplines, self-exclusion schemes and free treatment routes are listed in the helplines directory.