To examine cross-national changes in frequent adolescent cannabis use (40+ times consumed over life-time at age 15) over time and relate these trends to societal wealth, family affluence and gender.Data from three cycles (2002, 2006, 2010) of the Health Behaviour in School-aged Children (HBSC) Study were used for cross-sectional and trend analyses of adolescent cannabis use.Representative surveys in 30 European and North American countries.A total of 160 606 15-year-old students.Respondents' life-time cannabis use, demographics, family affluence (FAS) and frequency of peer contacts were measured individually. Indicators of wealth (gross domestic product per capita, GDP) and perceived availability of cannabis were obtained from national public data bases.The frequency of life-time cannabis use decreased over time among adolescents in Europe and North America, particularly in western European countries and the United States (relative risk (RR) = 0.86: confidence interval (CI) 0.79-0.93). This trend was not observed consistently in rapidly developing countries in eastern, central and southern Europe. Over time (2002-10), cannabis use became: (i) less characteristic of high GDP countries in contrast to lower GDP countries (RR = 0.74: CI 0.57-0.95); (ii) less characteristic of youth from high FAS families in contrast to youth from low FAS families (RR = 0.83: CI 0.72-0.96); and (iii) characterized by an increasing gender gap, i.e. consumption was higher among males (RR 1.26: CI 1.04-1.53). Perceived availability of cannabis and peer contacts remained strong predictors of frequent cannabis use.Among 30 European and North American countries, cannabis use appears to have 'trickled down' over time, with developing countries taking on the former (heavier) use pattern of richer countries, and less affluent youth taking on the former (heavier) use pattern of more affluent youth. Cannabis use continues to be more common among adolescent males than females.