Objective: The study examined the relationship between earlier adolescent marijuana use and later adolescent behavioral problems. Method: A community-based sample of Colombian adolescents was interviewed in 1995-1996 and 1997-1998. The time 2 (T-2) sample consisted of 1,151 males and 1,075 females. The psychosocial measures assessed adolescent problem behavior, the peer and sibling social network, and ecological/environmental stress and cultural domains. Logistic regression analyses included controls on demographic and time 1(T-1) dependent measures. Results: The findings suggest that T, adolescent marijuana use was associated with increased risks for T-2 adolescent difficulty at work or school, violent experiences, peer marijuana use, and sibling marijuana problems. Conclusions: This study provides important evidence in this cohort of the specific relationship between T-1 adolescent marijuana use and T-2 adolescent problem behavior in a society in which drug use, crime, violence, and low educational attainment are pervasive. Similar findings have been shown in previous research with U.S. adolescents. The findings suggest that early adolescent marijuana use is associated with an increase in problem behavior during later adolescence.