Eighth, ninth and tenth graders (N = 281) from two schools in suburban New York participated in a study to test the efficacy of a 10-session comprehensive psychosocial smoking prevention program. Schools were randomly assigned to experimental and control conditions. All students were given a pretest, posttest and three-month follow-up which assessed smoking behavior as well as knowledge and personality variables. There were significantly fewer new "smokers" in the experimental school than in the control school at both the initial posttest (p less than .01) and the three-month followup (p less than .05) along with significantly greater changes on some of the knowledge and personality variables. Overall, these results support the use of a smoking prevention strategy that focuses on the main psychosocial factors promoting the onset of cigarette smoking by teaching students basic coping skills.