Federal Data: Teen, Young Adult Marijuana Use Keeps Falling
On July 27, 2026, the Substance Abuse and Mental Health Services Administration published the 2025 National Survey on Drug Use and Health, and the timing alone is notable. It's the earliest release in the survey's history, built from more than 60,000 self-reported responses collected from Americans aged 12 and up across the country. Buried in that dataset is a finding that should reshape how lawmakers and parents talk about cannabis: teen and young adult marijuana use keeps falling, even as the map of legal states keeps filling in.
That trend runs directly counter to a warning that's been repeated in statehouses for more than a decade -- that legalizing cannabis for adults would inevitably funnel it toward kids. The 2025 data doesn't support that story. It suggests something closer to the opposite: as legal, regulated markets have expanded, adolescent use has kept sliding downward. What follows are the actual numbers behind that finding, what federal officials and advocacy groups are saying about it, and why a survey that nearly disappeared last year matters more than usual this time around.
What the 2025 NSDUH Actually Found
Photo by Brett Jordan via Unsplash.
The NSDUH is the federal government's flagship instrument for tracking substance use trends, and it carries weight precisely because of its scale and consistency. This edition drew on more than 60,000 self-reported responses, gathered through the same methodology SAMHSA has refined over decades, which is what allows researchers to compare 2025 figures against benchmarks stretching back to the early 2010s without worrying too much about apples-to-oranges problems.
Christopher D. Carroll, SAMHSA's Principal Deputy Assistant Secretary, framed this year's rollout as historic in its own right, calling it the earliest NSDUH release the survey has ever had. Beyond the scheduling milestone, Carroll pointed to a broader pattern: adolescents aged 12 to 17 and young adults aged 18 to 25 have shown downward or stable trends across most behavioral health indicators over the past five years, not just cannabis specifically.
The marijuana-specific numbers are the headline, though. Past-year use among 12-to-17-year-olds now sits at 8.7 percent, down sharply from 14.2 percent in 2011 -- the year before Colorado and Washington became the first states to legalize recreational cannabis for adults. Past-month use among that same age group fell from 6.1 percent, or roughly 1.6 million people, in 2021, to 5.0 percent, or about 1.3 million people, in 2025. Four years, a meaningful drop, and it happened during a stretch when the number of legal adult-use states roughly doubled.
The Numbers Behind the Decline
Smoking remains the dominant method of marijuana consumption among U.S. users in 2025 at nearly 75%, followed by eating/drinking and vaping, while dabbing is the least common method at just 14.5%.
NORML crunched the year-over-year figures and found that past-month teen marijuana use fell 22 percent between 2022 and 2025, while past-year use among the same group dropped 25 percent. Young adults aged 18 to 25 saw a smaller but still real decline, with past-month use down 10 percent over that same three-year window.
Compare that to adults 26 and older, where the trend line points the other way. Self-reported use in this group has been climbing in recent years, with the increase most pronounced among older adults -- people well past the age where curiosity or peer pressure typically drives first use. That's a meaningful contrast: the population gaining the most legal access on paper (adults, generally) is the one actually increasing consumption, while the population legalization critics warned about most (minors) is using less.
The survey also breaks down how people are consuming cannabis these days. Smoking remains the dominant method by a wide margin, reported by 74.8 percent of past-year users, or about 46.1 million people. Eating or drinking cannabis products came in second at 49.2 percent, roughly 30.3 million people, followed by vaping at 39.2 percent (24.2 million people) and dabbing at 14.5 percent (8.9 million people). Many respondents reported using more than one method, which is why these figures add up to more than 100 percent.
Put together, the generational split tells a more specific story than legalization increases use or legalization decreases use. It looks more like legalization is reshaping who uses cannabis and how, pulling in adults -- especially older adults trying regulated products for the first time -- without pulling in new teenage users the way prohibition-era predictions assumed it would.
A Survey That Almost Didn't Happen
Photo by Vitaly Gariev via Unsplash.
This year's on-time release is worth dwelling on because it wasn't guaranteed. Last year, the entire 17-member NSDUH team was laid off as part of Trump administration staffing cuts, and that move genuinely put the survey's future in doubt. Losing the people who design, field, and process a survey of this scale isn't a minor administrative hiccup -- it's the kind of disruption that can delay or degrade federal data collection for years.
Against that backdrop, an early and complete 2025 release isn't just routine bureaucratic continuity. It's a small but real signal that the survey infrastructure got rebuilt fast enough to keep the data pipeline intact, which matters for researchers, state health departments, and advocacy groups on all sides of the cannabis debate who rely on NSDUH as a common reference point.
It also isn't the only data point telling this story. The University of Michigan's Monitoring the Future survey, released in December 2025 and run independently of SAMHSA, found youth marijuana use holding essentially stable: 8 percent of 8th graders, 16 percent of 10th graders, and 26 percent of 12th graders reported past-year use. No surge, no spike tied to newly legal states -- just flat numbers consistent with the NSDUH decline.
Two federally connected but methodologically separate surveys, run by different institutions with different sampling approaches, are now converging on the same basic conclusion. When independent datasets agree this closely, it's harder to write off the finding as a fluke of survey design or a one-year statistical blip.
What This Means for the Legalization Debate
Photo by Robert So via Pexels.
Paul Armentano, NORML's Deputy Director, put the political implication plainly: alarmist claims tying legalization to rising teen use simply aren't backed by the government's own data. That's a pointed statement given how often that exact claim has been used in state legislative hearings and ballot measure campaigns over the past fifteen years.
Armentano's broader argument is that these findings should give lawmakers confidence that cannabis can be legally regulated in a way that doesn't compromise youth safety -- and arguably does a better job than prohibition did, since age-gated, licensed retail may be harder for minors to access than an unregulated street market ever was. That's a claim regulators in newer legal states will likely lean on as they defend existing frameworks against rollback efforts.
State-level research backs this up. A Marijuana Policy Project analysis found youth marijuana use declined in 19 of the 21 states that had legalized adult-use cannabis at the time of the study. That's not a razor-thin trend -- it's a strong majority of legal states showing the same downward direction the federal data shows nationally.
Federal and state-level numbers are lining up here, and that convergence chips away at one of the most durable arguments prohibition advocates have used for over a decade. It's worth remembering, though, that cannabis law is still a patchwork -- rules on possession, purchase age, home cultivation, and public consumption vary considerably from state to state and country to country. Anyone making decisions based on this data should check current regulations in their own jurisdiction rather than assume national trends translate into local legality.
Ten-plus years of expanding legal cannabis access have now run parallel to a steady decline in teen use, not the surge that prohibition-era arguments insisted was inevitable. That's not a small footnote -- it's a decade-long natural experiment across dozens of states, and the data keeps pointing the same direction.
The more interesting open question at this point isn't about teenagers, it's about older adults. Why does use keep climbing in that group specifically? The likeliest explanation is straightforward: regulated retail access is finally reaching people who spent decades avoiding cannabis entirely, whether out of legal risk, stigma, or simply never having a legitimate storefront to walk into. A 55-year-old checking out a licensed dispensary for chronic pain or sleep issues is a very different phenomenon than a 15-year-old buying from a friend at school, and the numbers reflect that distinction clearly.
There's one more thing worth watching going forward, separate from whatever the next NSDUH finds: whether it gets released at all, and on schedule. After last year's staffing cuts gutted the entire survey team, this year's early rollout was as much a story about institutional resilience as it was about the numbers themselves. The data is only useful if it keeps showing up, year after year, built on the same rigorous methodology. That continuity is worth defending regardless of where anyone lands on legalization itself.
