The research format consisted of a pre-and post-test experimental and control group design that followed 619 students from late primary school until high school in grade 10. Longitudinal and cross sectional analysis of data for major drug categories: alcohol, tobacco, analgesics, and marijuana is presented to identify trends and the relationship between attitudes and behavior. In addition, regression analysis focuses on the predictive power of attitudes and specific drug use behavior for subsequent drug use and the ability to resist peer pressures. Results indicate that an early education program based on psychosocial principles can produce positive changes in drug taking behavior, although booster session throughout the middle years are required. 21 references, 5 tables, and 3 figures. (Author abstract modified)