Results from a longitudinal experiment to curb drug use during junior high indicate that education programs based on a social-influence model can prevent or reduce young adolescents' use of cigarettes and marijuana. This multi-site experiment involved the entire seventh-grade cohort of 30 junior high schools drawn from eight urban, suburban, and rural communities in California and Oregon. Implemented between 1984 and 1986, the curriculum's impact was assessed at 3-, 12-, and 15-month follow-ups. The program, which had positive results for both low- and high-risk students, was equally successful in schools with high and low minority enrollment. However, the program did not help previously confirmed smokers and its effects on adolescent drinking were short-lived.