iGaming Journalist & Crypto Casino Analyst
Ask two credible organisations how many Americans gamble and you will get two very different answers. In August 2026, Gallup reported that 45% of U.S. adults had gambled in the past year. The American Gaming Association's most recent attitudes research puts past-year participation at 57%. Both are national, both are professionally conducted, and both are the numbers that end up in headlines, policy submissions and university papers.
This article assembles the U.S. gambling demographic data into one reference set: participation by activity, by age, by income and by gender, alongside the revenue figures that show where the money actually lands. It also explains why the headline participation numbers disagree by twelve percentage points, because anyone citing this data needs to understand what they are citing.
Key Findings
Gallup's 2026 poll puts past-year U.S. gambling participation at 45%, down from 64% in 2016. Lottery play fell from 49% to 31% and in-person casino visits from 26% to 14% over the same decade. The American Gaming Association records 57% participation and a record $78.72 billion in commercial gaming revenue for 2025, up 9.2%. Sports betting skews young and affluent: 29% of bettors earn over $100,000. Roughly 8% of adults report at least one problem-gambling indicator.
The Headline Number Depends on Who Is Asking
The single most misused statistic in this field is "the percentage of Americans who gamble." There is no one number, and the gap between the available figures is not noise.
Gallup's poll asks respondents whether they have taken part in each of a list of specific activities in the past twelve months, then aggregates. Its August 2026 release put the combined figure at 45%, a nineteen-point fall from the 64% Gallup recorded in 2016. The American Gaming Association's survey work, which asks about gambling participation in a gaming-industry context, records 57% past-year participation, with 30% having visited a physical casino and 21% having placed a sports bet.
Three structural differences explain most of the divergence:
- Question framing. Gallup enumerates activities one at a time; a respondent who bought a single scratch card in March may not recall it when prompted about "buying a state lottery ticket."
- Social desirability. Gallup found that the share of Americans calling gambling "morally acceptable" fell from 67% in 2016 to 57%. When an activity becomes less socially approved, self-reported participation falls faster than actual participation.
- Sample and mode. Telephone-based polling reaches a different population than online panels, and online panels reach people who are, by definition, comfortable online — where most new gambling now happens.
The practical guidance for anyone citing these figures: use Gallup for trend direction over time, because its methodology has been consistent for decades, and use industry and regulator data for absolute volume, because revenue is measured rather than remembered.
Participation by Activity: The Decade of Change
Gallup's activity-level breakdown is the most useful longitudinal series available, because the same questions have been asked in the same way since the 1990s. The 2016-to-2026 comparison covers exactly the period in which the Supreme Court struck down PASPA and legal sports betting expanded to most of the country.
| Activity | 2016 | 2026 | Change |
|---|---|---|---|
| Any gambling (past 12 months) | 64% | 45% | -19 pts |
| Bought a state lottery ticket | 49% | 31% | -18 pts |
| Visited a casino | 26% | 14% | -12 pts |
| Participated in an office sports pool | 15% | 7% | -8 pts |
| Gambled online for money | — | 4% | Flat |
| Say gambling is morally acceptable | 67% | 57% | -10 pts |
The counter-intuitive result is worth stating plainly: reported participation fell across every traditional category during the single largest legal expansion of gambling in American history. Online gambling is the only category that did not decline, and it is reported by just 4% of adults — a figure impossible to reconcile with the measured revenue in regulated online markets.
The most defensible reading is that the activity has concentrated. Fewer people gamble, but those who do gamble more often and spend more, and a growing share of that spend has moved to phones where it is less memorable as a discrete event. A weekly $2 lottery ticket is an occasion; forty in-play micro-bets during an NFL game may not register as forty acts of gambling in a survey respondent's memory.
Age: Three Different Gambling Populations
Aggregate participation figures hide the fact that the U.S. does not have one gambling population — it has at least three, each attached to a different product.
| Age group | Highest-participation product | Notable figure |
|---|---|---|
| 21-34 | Sports betting and online | ~32% placed a sports wager in past 12 months |
| 35-54 | Land-based casino | Highest frequency of casino visits |
| 55+ | Lottery | Highest lottery participation rates |
Adults aged 21 to 34 report the highest sports betting participation of any cohort, at roughly 32% in the past year, according to AGA research. Adults 35 to 54 make the most frequent land-based casino visits. Adults 55 and over participate most heavily in lottery games.
This matters for anyone modelling the industry's future. The lottery-heavy cohort is ageing out; the sports-betting cohort is entering peak earning years. If product preference is generational rather than a function of life stage, the revenue mix will keep shifting toward mobile and sports-linked products for at least another two decades. If it is a life-stage effect, today's 25-year-old sports bettor becomes tomorrow's casino visitor. The available longitudinal data is not yet long enough to settle the question, and any source claiming otherwise is extrapolating.
The National Council on Problem Gambling's NGAGE research has also drawn attention to pre-21 exposure, finding widespread gambling participation before the legal age — predominantly through informal betting, fantasy contests and social casino games rather than regulated operators.
Income and Education
The stereotype of gambling as a regressive activity concentrated among low-income households holds for lottery products and breaks down badly for sports betting.
AGA research finds that 29% of sports bettors report household income above $100,000 — almost double the proportion in the general adult population. Sports betting participation correlates positively with income, education and sports fandom. Lottery participation runs in the opposite direction, with the heaviest per-capita spending consistently found in lower-income ZIP codes across state-level studies.
Anyone writing about "gambling and inequality" therefore needs to specify the product. The regressivity finding is real, robust and about lotteries and, to a lesser degree, electronic gaming machines. Applying it wholesale to the sports betting market misstates the demographics by a wide margin. Readers looking for how the mechanics of each product differ can start with our guides hub.
Gender
Gender differences appear most sharply in risk statistics rather than participation. Prevalence studies consistently find gambling disorder roughly twice as common among men as among women, though multiple recent surveys suggest the gap is narrowing. Product choice diverges too: sports betting and poker skew heavily male, while bingo and certain slot categories skew female or close to even.
The narrowing of the disorder gap is one of the more important trends in the literature, and it tracks the shift to mobile. Products that required entering a physical, male-coded space had a built-in demographic filter. An app does not.
Where the Money Actually Is
Survey data measures what people say. Regulator filings measure what operators booked. The second is far more reliable for sizing the market.
| Metric (U.S. commercial gaming, 2025) | Value | Year-on-year |
|---|---|---|
| Total gross gaming revenue | $78.72 billion | +9.2% |
| Traditional gaming (casino floor) | $50.94 billion | +2.3% |
| Sports betting revenue | $16.96 billion | +22.8% |
| Sports betting handle | $166.94 billion | +11.0% |
| Gaming tax paid to states | $18.09 billion | +15.1% |
| Markets setting annual records | 34 of 38 states plus DC | — |
Every one of the 38 commercial gaming markets grew in 2025, and Washington DC posted the largest percentage increase in consumer spend at 75.9%. Note the divergence in growth rates: the casino floor grew 2.3% while sports betting revenue grew 22.8%. That is the same generational split visible in the participation data, expressed in dollars.
The tax figure is the one most often cited by legislators. At $18.09 billion, state and local gaming tax receipts grew faster than revenue itself, reflecting both new market launches and higher rates in established states. Anyone comparing state tax yields should be careful to distinguish gross gaming revenue taxes from promotional-deduction-adjusted taxable revenue, which differ substantially by state. Our US sports betting hub tracks the state-level picture.
Risk and Harm Indicators
The NCPG's NGAGE survey series is the most consistent national instrument for gambling risk indicators. Its recent findings:
- Roughly 8% of U.S. adults — close to 20 million people — reported at least one indicator of potentially problematic gambling "many times" in the past year.
- That is down from 11% in 2021, but above the 7% recorded before widespread legal sports betting.
- Among sports bettors, 17% of traditional bettors and 19% of online bettors reported at least one such indicator.
- An estimated 2.5 million adults meet criteria consistent with gambling disorder, with a further 5 to 8 million showing some problematic behaviour.
- Parlay betting rose from 17% of sports bettors in 2018 to 30% in 2024 — a shift toward higher-hold, higher-variance products.
The parlay figure deserves attention from researchers. Parlays carry substantially higher operator hold than single-game wagers, so a shift in bet mix raises revenue per dollar staked without any increase in handle. Any analysis of rising sportsbook revenue that does not control for bet mix will attribute to volume what is partly attributable to product design.
How the Legal Map Shapes the Demographics
National averages flatten out enormous state-level variation, and that variation is itself demographic data. A 24-year-old in New Jersey has legal access to online casino games, online poker and mobile sports betting from the same phone. A 24-year-old in Georgia has no legal online option at all. Both appear in the same national participation figure.
Three consequences follow, and each one is a common source of error in published analysis:
- Participation gaps are partly access gaps. Differences in reported gambling rates between states reflect what is legally available as much as what residents prefer. Cross-state behavioural comparisons that do not control for market status are measuring regulation, not culture.
- Handle per adult is the better comparator. Because regulated handle is measured rather than self-reported, per-adult handle in states with mature markets is the most reliable indicator of real engagement intensity.
- Launch timing distorts growth rates. A state in its first full year of legal betting will post growth percentages that a mature market cannot match. The 75.9% consumer-spend growth recorded in Washington DC in 2025 is a market-structure story, not a demand story.
The same caution applies to online casino data. Only a handful of states have licensed iGaming, so national "online casino participation" figures are averages across a country where most residents cannot legally participate. Readers comparing regulated options can review our online casino reference for how state availability differs by product.
How to Use This Data Responsibly
For journalists, students and analysts citing these figures, four rules prevent the most common misstatements:
- Always pair a participation percentage with its source and year. "45% of Americans gamble" without "Gallup, 2026" is not a usable citation, because a different and equally valid source says 57%.
- Never mix survey participation with revenue growth in the same causal sentence. They move in opposite directions right now and measure different things.
- Specify the product when discussing income effects. Lottery and sports betting have close to opposite income profiles.
- Treat problem-gambling prevalence figures as ranges. Instruments differ, thresholds differ, and the difference between "at least one indicator" and "meets diagnostic criteria" is a factor of roughly eight.
Methodology
All figures in this article are drawn from published primary sources: Gallup's national polling series, American Gaming Association research and its 2025 commercial gaming revenue report, and the National Council on Problem Gambling's NGAGE survey programme. No figures have been estimated, modelled or interpolated by DeucesCracked.
Where two sources disagree, both figures are presented with their provenance rather than averaged. Survey-based participation figures carry sampling error typically in the range of ±2 to ±4 percentage points at the national level, and self-reported gambling participation is subject to known under-reporting bias. Revenue and tax figures are compiled from state regulator filings and should be treated as materially more precise than survey data.
Commercial gaming revenue excludes tribal gaming, which is reported separately and adds substantially to the national total. Comparisons across states are complicated by differing definitions of taxable revenue, particularly around promotional credit deductions.
Frequently Asked Questions
What percentage of Americans gamble?
Between 45% and 57% depending on the survey. Gallup's 2026 poll reports 45% past-year participation; American Gaming Association research reports 57%. The difference reflects question wording, survey mode and social-desirability bias rather than a factual dispute.
Is gambling participation in the U.S. rising or falling?
Self-reported participation is falling — Gallup shows 64% in 2016 against 45% in 2026 — while measured revenue is rising, reaching a record $78.72 billion in 2025. The most likely explanation is concentration: fewer participants, higher spend per participant, and a shift to mobile activity that respondents recall less reliably.
Which age group gambles the most?
It depends on the product. Adults 21-34 lead sports betting at roughly 32% past-year participation, adults 35-54 visit land-based casinos most frequently, and adults 55+ have the highest lottery participation.
Do high earners bet on sports?
Yes, disproportionately. Some 29% of sports bettors have household income above $100,000, roughly double the share in the general population. Sports betting demographics differ sharply from lottery demographics, which skew lower-income.
How many Americans have a gambling problem?
NCPG research indicates about 2.5 million adults show behaviour consistent with gambling disorder, with 5 to 8 million more exhibiting some problematic behaviour. Around 8% of adults report at least one problem-gambling indicator "many times" in the past year, down from 11% in 2021.
Sources
- Gallup — Fewer U.S. Adults Say They Gamble (2026)
- American Gaming Association — American Attitudes Toward Gaming
- AGA — Commercial Gaming Revenue Hits $78.7 Billion in 2025
- National Council on Problem Gambling — NGAGE 3.0
Cite This Article
If you use data from this article, please link back to https://www.deucescracked.com/blog/us-gambling-demographics-data-study. For related reference material see our casino hub and research blog.
Join the Conversation
Be respectful. No spam. Strategy discussion welcome.