Kuntsche et al1 are to be applauded. The Health Behavior in School-Aged Children Study is a very ambitious and important ongoing international collaboration that these investigators and their many colleagues have been conducting, one that will continue to offer increased understanding of the epidemiology and etiology of substance use. This repeated cross-sectional (time-lag) project clearly shows the advantages of international collaboration, providing needed insight into robust vs more culture-specific phenomena related to the causes, covariates, and consequences of adolescent drug use. Regarding the current article, this simple but elegant analysis tells an important story: it confirms that marijuana use among adolescents has declined in recent years in several European and North American countries2; it confirms the relatively strong correlation between marijuana use and evenings out, consistent with what has been found based on US national samples3,4; and most importantly, the findings document that cannabis use and evenings out covary across a 4-year period (2002-2006) such that the decline in cannabis use is associated with a decline in evenings out. Even as the level of cannabis use and evenings out shifted over time, the relationship between the two was steady, confirming the general conclusion that even as substance use and associated risk factors vary in level over time, the size of the relationships tends to remain invariant.5