In the study's first phase, participants were randomly assigned by groups to the experimental (200 participants) and control (235 participants) conditions. The number of participants was subsequently reduced by 38 percent due to some participants' ceasing to be members of the organization. The experimental group received the LSEP, which used a cognitive behavioral approach. All participants were tested prior to the beginning of LSEP, immediately upon its completion, and approximately 1 month later, at the end of new employee training. A follow-up study 3 years later gathered information by means of structured telephone interviews. The Drug Attitude Scale was mailed to participants who answered the structured interviews. A total of 226 questionnaires returned. Data were analyzed using a two-way multivariate analysis of variance. Results revealed that the LSEP experience had little effect on increasing knowledge of policy toward drug and alcohol use in either the short-term or the long-term. However, LSEP had an immediate and long-lasting effect on attitudes toward marijuana, although it had only a short-term effect on attitudes toward alcohol. Moreover, LSEP had no significant effect on actual alcohol consumption, but very few of the soldiers in both groups abstained from drinking. Furthermore, compared to the control group, significantly more of the LSEP participants reported having used marijuana and hashish during the last 12 months. Findings suggested that a life-skills approach is probably better suited for secondary prevention for adults than for primary prevention. Tables and 45 references