Programs which have been somewhat effective in reducing the rates of onset of regular tobacco use have featured such components as peer pressure resistance training, correction of normative expectations, inoculation against mass media messages, information about parental influences, information about consequences of use, public commitments, or peer opinion leaders. Most of these programs have been implemented by highly trained personnel. This study examined the effectiveness of a tobacco and alcohol prevention program delivered to sixth and seventh grade students by minimally trained classroom teachers. The program focused on: (1) teaching students to identify and resist peer influences; (2) information about short- and long-term consequences of tobacco and alcohol use; (3) the correction of normative expectations; and (4) the establishment of conservative intentions regarding tobacco and alcohol use. Two cohorts of students were pretested and, subsequent to delivery of the program, tracked longitudinally; the first cohort (N=1,221) was followed for 4 years, the second (N=1,707) was followed for 3 years. Results indicated that the program reduced the onset and prevalence of tobacco use but not of alcohol use. The effects for tobacco were differentially related to school district in which the program was delivered, sex of student, and ethnicity suggesting that prevention program content is only one variable that may affect the effectiveness of prevention interventions. (Author/NB)