Harmful effects of cannabis use in youth scale with frequency. In the context of approaching legalization in Canada, understanding the distinctions between youth who change and who maintain their cannabis use is essential for designing interventions and policy. A substantial number of characteristics may play a role. This study assessed whether and how youth who change their cannabis use differ from those who do not.Cannabis use was reported by 8375 Canadian high school students participating in the COMPASS study. GEE multiple logistic regressions were then used to establish impact of 13 baseline demographic and behavioural exposure variables on the likelihood of membership in four frequency change groups at follow-up a year later: reduction, cessation, escalation, and maintenance.Groups were found to differ in several ways. Cessation group members (19.4% of users) were less likely to binge drink (OR 0.82), vape (OR 0.82), or attend class without completing homework (OR 0.72) than maintainers. Students who reduced their use (14.6% of users) were more likely to binge drink (OR 1.36), smoke (OR 1.20), and skip class (OR 1.21). Those who escalated (29.5% of users) were more likely to be male (OR 1.35) and to vape (OR 1.22).Students who change their cannabis use differ in several demographic and behavioural characteristics. The results raise further concerns about the impact of e-cigarettes and the role of poly-substance use in high-risk trajectories. Distinct classes of cannabis users, essential for policy and intervention development, can be identified in high school populations.