12-year Washington young-adult survey: use rising among 21–25, falling among 18–20, researcher says

Dr. Jason Kilmer, a professor of psychiatry and behavioral sciences at the University of Washington, told the Liquor and Cannabis Board caucus on May 19, 2026, that 12 years of the Washington Young Adult Health Survey show diverging trends by age: use and frequency have increased for 21–25 year olds while decreasing for 18–20 year olds.

The survey, which Kilmer said has collected responses from 21,494 unique individuals since 2014, is a convenient sample recruited by a mix of Department of Licensing mailings and online advertising and is weighted to improve representativeness. "We've been able to move into 12 year territory — that's remarkable," Kilmer said, noting that the team applies census-based weights so the weighted results align closely with unweighted findings.

Why it matters: the survey isolates the 18–25 age band that national surveys often group more broadly. Kilmer told the board that this focus allows the team to compare those with legal access to retail cannabis (21–25) with those who are under 21, and it helps trace how retail access and shifting norms relate to use.

Key findings Kilmer highlighted: overall any past-year cannabis use is no longer showing a statistically significant linear trend across all 18–25 year olds because the age groups moved in opposite directions — a significant decrease for 18–20 year olds and a significant increase among 21–25 year olds. Frequency measures are similar: monthly and weekly use trends are increasing overall but are driven by the older cohort, while the younger cohort shows declines. "Those with legal access to the retail market are the ones driving that significant increasing trend over time," Kilmer said.

Kilmer also described changes in sources and perceptions. Among 18–20 year olds, endorsements for getting cannabis from friends, medical-card holders, dispensaries and parties declined, while purchasing through intermediaries, parents with permission and even stealing rose. For 21–25 year olds, most sources declined except buying from retail stores, which increased. Kilmer called attention to a pronounced misperception: although a majority of respondents do not use, many young adults overestimate how often their peers use — for one cohort he noted, 17% used at least weekly while 62% believed the typical person their age used weekly or more.

On driving after use, Kilmer said self-reported driving within three hours of cannabis use has decreased compared with early cohorts, but remains notable: "About 30 percent in cohort 12 of young adults that reported past 30 day use also reported driving within 3 hours of use," he said, referencing Washington Traffic Safety Commission data as context for continued concern.

Other substance trends and perceptions: Kilmer reported declines in multiple other substances (including alcohol, cigarettes, prescription pain reliever misuse and heroin). He also reported a recent reversal in perceived harm for some cannabis measures: perceived physical and psychological risks that had been declining now show newly significant increases — a pattern Kilmer said has correlated with later short-term changes in use in national data.

Board members asked clarifying questions during and after the presentation. When a board member raised the lower participation in the 2025 cohort, Kilmer explained that recruitment opened late because of funding delays, leading to a shorter window for enrollment, but said the weighted results still matched the unweighted data. On potency, Kilmer described a new survey feature that lets respondents report potency as a percentage or milligrams based on preferred product type; he said responses show a bimodal distribution with many reporting mid-range percentages and a substantial group reporting higher percentages.

Kilmer told the board the team submitted the longitudinal reports to the Division of Behavioral Health and Recovery and will share PDFs with staff; he said the survey will enter a thirteenth year of data collection next year. Chair Vallendorf and board members thanked Kilmer for the presentation.

The board did not take formal action on the presentation; Kilmer's report was informational and staff were encouraged to follow up individually with him for additional detail.