OBJECTIVES The purpose of this study was to implement and evaluate the effectiveness of a long-term tobacco use prevention program for junior high school students that used college undergraduate change agents and telephone boosters. METHODS A psychosocial intervention combining refusal skills training, contingency management, and other tobacco use prevention methodologies such as telephone and mail boosters was implemented in 11 junior high schools in San Diego County, California. Eleven other junior high schools served as controls. Of the 2668 participants, 57% were White/non-Hispanic, 24% were Hispanic, and 19% were of other racial/ethnic groups. College undergraduates served as change agents for both the classroom and booster interventions, the latter of which was delivered in the third (ninth-grade) year of the program. RESULTS At the end of the third year, the prevalence of tobacco use within the past month was 14.2% among the intervention students and 22.5% among the controls, yielding an odds ratio of 0.71 for analysis at the school level. CONCLUSIONS Both college undergraduate change agents and direct one-to-one telephone interventions appear to provide cost-effective tobacco-related behavior modification.