Cannabis Statistics 2026: The Definitive, Sourced Data Hub

How many Americans use cannabis, where it's legal, how big the legal market is, and what the data actually says — every number cited to a real, named source.

By Justin Park · ~14 min read · Updated 2026-06-22

Take the 20-second finder

If you've ever tried to pin down a real cannabis number, you know the problem: the internet is flooded with stats that have no source, contradict each other, or quietly recycle a figure from 2016. This page is our answer to that. It's a living reference, updated for 2026, where every statistic is tied to a named, authoritative source — SAMHSA, the CDC, NIDA, Gallup, Pew, NHTSA, Monitoring the Future, and established market analysts — so you can cite it, quote it, or just settle an argument with confidence.

A quick note on how to read this honestly. Survey data and government reports run on a lag, so the most recent hard federal numbers often describe 2023 or 2024 even in 2026 — we label the year every time. We also resist the urge to round everything in cannabis's favor. Use is rising among adults and falling among teens; the legal market is enormous and also under real financial strain; public support is historically high and recently dipped. All of that is true at once, and a good data hub says so.

Below you'll find usage and prevalence, demographics, the legal map, market size and tax revenue, medical patient counts, workplace testing, impaired-driving research, youth trends, the CBD market, and public opinion — with tables you're welcome to reproduce as long as you keep the sourcing intact.

The short version

  • <strong>43.6 million Americans</strong> (15.4% of people 12 and older) used cannabis in the past month, according to SAMHSA's <strong>2023 National Survey on Drug Use and Health</strong> — the most recent federal figure available.
  • As of 2026, <strong>24 states plus Washington, D.C.</strong> have legalized adult-use (recreational) cannabis, and roughly <strong>38–39 states plus D.C.</strong> have a comprehensive medical cannabis program.
  • Legal U.S. cannabis retail sales topped <strong>$30 billion in 2024</strong> (NORML analysis) and are estimated around <strong>$33–34 billion in 2025</strong> — and states have collected over <strong>$20 billion in cumulative cannabis tax revenue</strong> since 2014.
  • <strong>64% of U.S. adults</strong> said cannabis should be legal in <strong>Gallup's November 2025 poll</strong> — still a clear majority, though down from the 68–70% range of 2020–2024.
  • Teen use keeps <em>falling</em>: <strong>Monitoring the Future</strong> found past-year cannabis use among 12th graders at about <strong>26% in 2025</strong>, near a 30-year low, despite widespread legalization.
  • Among seriously or fatally injured road users in an NHTSA study (2019–2021), <strong>25.1% tested positive for active THC</strong> — the most common substance found, though detection alone does not prove impairment.

The 20-second finder

Not sure which is right for you?

Answer a few quick questions and we'll point you to the pick that fits — from this guide's lineup.

Find your match

30-sec finder

Question 1 of 4

First things first — how do you want to feel?

The headline number: how many people use cannabis

In 2023, an estimated 43.6 million Americans aged 12 or older — 15.4% of that population — used marijuana in the past month, and roughly 61.8 million (about 21.8%) used it in the past year, according to SAMHSA's 2023 National Survey on Drug Use and Health (NSDUH). That past-month figure includes about 15.8 million people (5.6%) who reported vaping cannabis. NSDUH, run by the federal Substance Abuse and Mental Health Services Administration, is the largest and most-cited U.S. survey of self-reported substance use, which is why it anchors this page.

For context on the trend: past-month cannabis use has climbed steadily over the last two decades as legalization spread and stigma fell. The 2023 NSDUH past-month rate of 15.4% is among the highest the survey has recorded. The honest caveat is that all of this is self-reported, and reporting tends to rise as legal risk falls — so some of the increase reflects people being more willing to say yes, not only more people using.

Past-year cannabis use by age (2023)

Cannabis use is far from evenly distributed. Young adults aged 18 to 25 are consistently the heaviest-using group, while use among adolescents is comparatively low and falling (see the youth section below). The table reflects SAMHSA's 2023 NSDUH past-month estimates, the cleanest age breakdown the federal survey publishes.

Age groupPast-month use (2023)Approx. number of people
12–17 (adolescents)~6.5%~1.7 million
18–25 (young adults)25.2%~8.6 million
26 or older15.0%~33.5 million
12 and older (total)15.4%43.6 million

Source: SAMHSA, 2023 National Survey on Drug Use and Health (NSDUH). Note that adults 26+ are the largest single block of users in raw numbers simply because the group is so big, even though young adults use at a higher rate.

Registered medical cannabis patients

The number of registered medical cannabis patients in the U.S. surpassed 5 million in 2023, and a December 2025 White House statement referenced more than 6 million registered patients nationwide. The true count is genuinely uncertain, because several states with adult-use markets dropped or never maintained mandatory patient registries — once anyone over 21 can buy legally, many former medical patients simply stop renewing cards, which makes the registered total undercount medical use in some states even as it grows in others.

On average, where registries exist, somewhere around 1–2% of a state's population enrolls in the medical program. Patients most commonly register for chronic pain, with PTSD, anxiety, cancer-related symptoms, and seizure disorders also prominent across state qualifying-condition lists. As always with cannabis and health: this is general information, not medical advice, and qualifying conditions vary state by state.

Cannabis-impaired driving: what the data shows

This is the area where honest framing matters most. In an NHTSA study of seriously or fatally injured road users at seven Level 1 trauma centers (data collected September 2019–July 2021), 25.1% tested positive for active THC — making cannabinoids the single most frequently detected substance, ahead of alcohol at 23.1%. Overall, 55.8% of those road users tested positive for at least one potentially impairing substance.

But detection is not the same as impairment. NHTSA itself cautions that finding THC in a person's system does not establish that cannabis caused or contributed to a crash, because THC can remain detectable long after any impairing effect has worn off, and many of these cases involved multiple substances. Earlier NHTSA roadside surveys did document a real rise in drivers testing positive for THC — from 8.6% of weekend nighttime drivers in 2007 to 12.6% in 2013–2014 — so the prevalence trend is upward even if causation is hard to isolate. The responsible takeaway, and the one we'd want any cannabis user to hear: driving impaired is dangerous and the research on roadside prevalence is rising, even though THC tests don't cleanly measure impairment the way a breathalyzer measures alcohol.

Youth use is falling — the counterintuitive headline

Here's the finding that most surprises people who assume legalization floods schools with weed: teen cannabis use has been falling and now sits at or near a 30-year low. According to the University of Michigan's Monitoring the Future (MTF) survey — the long-running, NIDA-funded gold standard for adolescent drug-use data — past-year cannabis use among 12th graders was about 26% in 2025, with 8th and 10th graders far lower, and 8th-grade past-year use dropping another 16% year-over-year into 2025.

GradePast-year cannabis use (2025, approx.)Long-term trend
8th grade~6–7%Down sharply over the decade
10th grade~15–16%Lowest in three decades
12th grade~26%Near 30-year low

Source: Monitoring the Future, University of Michigan / National Institute on Drug Abuse (NIDA), 2024 and 2025 surveys. Researchers have repeatedly noted that the wave of state legalization since 2012 has not produced the predicted spike in youth use — if anything, the lines kept dropping, a point NIDA and MTF investigators make directly.

The CBD and hemp market

Cannabidiol (CBD) — the non-intoxicating cannabinoid that went mainstream after the 2018 Farm Bill legalized hemp federally — is its own large market sitting alongside THC cannabis. Statista's market outlook puts U.S. CBD product revenue at roughly $3 billion in 2025, while broader research firms estimate the global cannabidiol market in the $10 billion range for 2025, growing at double-digit annual rates. The spread between estimates is wide because firms define "CBD market" differently (oils only vs. all hemp-derived products), so treat any single CBD figure as an order-of-magnitude guide rather than a precise count.

The hemp side of the industry has also become legally and politically tangled, as intoxicating hemp-derived cannabinoids like delta-8 THC exploited the gap between hemp legality and cannabis prohibition — a regulatory loose end that lawmakers were actively trying to close heading into 2026.

Public opinion: support for legalization

A clear majority of Americans support legal cannabis, but the trend recently ticked down. Gallup's November 2025 poll found 64% of U.S. adults think cannabis should be legal — a strong majority, but below the 68–70% range Gallup recorded from 2020 through 2024. Gallup attributed the dip largely to a 13-point drop in support among Republicans (to 40%), while Democrats (85%) and independents (66%) held steady.

Pew Research Center frames the question differently and finds even broader acceptance of some legal access: in its early-2024 survey, about 88% of Americans said cannabis should be legal for medical use, recreational use, or both, with 57% favoring legalization for both purposes and just 11% saying it should not be legal at all. The long arc is unmistakable — support has roughly doubled since the early 2000s.

YearGallup: % saying cannabis should be legal
200031%
201358% (first majority)
201966%
202168%
202268%
202564%

Sources: Gallup (annual Crime and Consumption polls, including November 2025); Pew Research Center (2024).

How to use these numbers (and what to watch for)

A few habits will keep you out of trouble when citing cannabis data. Always name the source and the year — "15.4% of Americans 12+ used cannabis in the past month (SAMHSA, 2023 NSDUH)" is bulletproof; "most Americans use weed" is not. Distinguish detection from impairment in any driving or workplace stat. Treat market and CBD forecasts as ranges, not facts, because they're modeled and frequently revised. And remember the lag: in 2026, the freshest federal usage data still describes 2023.

If a number on this page ever looks stale to you, that's a feature — we date everything precisely so you can tell. This is general information for context and curiosity, not medical or legal advice, and cannabis laws and health guidance vary by state and change often.

Key terms

NSDUH
The National Survey on Drug Use and Health, run annually by SAMHSA — the largest U.S. self-report survey of substance use and the standard source for national prevalence figures.
Past-month vs. past-year use
Two standard survey windows. 'Past-month' (a.k.a. 'current use') counts anyone who used in the prior 30 days; 'past-year' is a broader 12-month window and yields larger numbers.
Adult-use (recreational)
Cannabis legalized for any adult, typically 21+, without a medical recommendation — distinct from medical-only programs.
Monitoring the Future (MTF)
A long-running, NIDA-funded University of Michigan survey of 8th, 10th, and 12th graders — the leading source for U.S. youth drug-use trends.
Section 280E
A federal tax provision that bars businesses 'trafficking' Schedule I/II substances from deducting normal expenses — a major reason legal cannabis operators face heavy effective tax rates.
Schedule I / Schedule III
Federal drug classifications. Cannabis has been Schedule I (no accepted medical use); a proposed move to Schedule III would loosen research restrictions and ease the 280E tax burden.

Questions, answered

How many Americans use cannabis?

About 43.6 million Americans aged 12 or older — 15.4% of that population — used cannabis in the past month, and roughly 61.8 million used it in the past year, according to SAMHSA's 2023 National Survey on Drug Use and Health, the most recent federal data available as of 2026.

How many states have legalized cannabis?

As of 2026, 24 states plus Washington, D.C. have legalized adult-use (recreational) cannabis, and roughly 38–39 states plus D.C. have a comprehensive medical cannabis program. D.C. allows possession but not commercial sales. Cannabis remains federally controlled, with a Schedule III rescheduling under review.

How big is the legal cannabis market?

Legal U.S. retail cannabis sales topped $30 billion in 2024 (NORML analysis of state data) and are estimated around $33–34 billion in 2025. States have collected over $20 billion in cumulative cannabis tax revenue since 2014, per the Marijuana Policy Project.

Is teen cannabis use going up because of legalization?

No — the data shows the opposite. Monitoring the Future found past-year cannabis use among 12th graders at about 26% in 2025, near a 30-year low, with younger grades far lower and still falling. Researchers note the predicted surge in youth use after legalization did not materialize.

What percentage of Americans support legalizing cannabis?

Gallup's November 2025 poll found 64% of U.S. adults support legalization — a majority, though down from the 68–70% range of 2020–2024. Pew Research found about 88% support legal cannabis for medical use, recreational use, or both, with 57% favoring both.

Does THC in a crash or drug test prove someone was impaired?

No. NHTSA found 25.1% of seriously or fatally injured road users tested positive for active THC (2019–2021 data), but the agency stresses that detection does not establish impairment, because THC can remain detectable long after any effect has worn off. The same caveat applies to workplace tests.