A quasi-experimental study was conducted to explore the efficacy of the program "Stay Away From Tobacco" (SAFT). Eleven classes consisting of 381 students in grades 7, 8, 10 and 11 were assigned by class to 3 groups (Intervention Group T with school teachers delivering the program, Intervention Group R with researchers delivering the program, and Control Group C). Data were collected at baseline, immediately after intervention and six month post-intervention. Self-reported smoking was the outcome measure. Program effect was assessed using the GLIMMIX and the MIXED methods. The 30-day smoking prevalence in group C increased from 4% at baseline to 10% at the six-month follow-up; while the same rate declined from 11% to 6% in group T, and from 9% to 1% in group R. For group T, the odds ratio (for 30-day smoking) and the regression coefficient (for indexed number of cigarettes smoked) assessing interactions between intervention and time was 0.20 (p<.001) and -0.1605 (p<.05) respectively. The same statistics for group R were 0.09 (p<.001) and -0.2406 (p<.01) respectively. There was a 19% decline in the predicted smoking rate from baseline to six-month follow-up in group T (11.5% vs. 9.3%) and a 26% decline in the same rate in group R (11.1% vs. 8.2%). The results from this pilot trial suggest that SAFT can reduce cigarette smoking among middle and high school students through its effect on improving these students' refusal skills and changing their perceived mental and physical values from smoking. A full scale evaluation is recommended.