The focus of the article is on school-based programs and primary prevention efforts that typically take place before more advanced alcohol and drug use is observed. Meta-analysis is conducted of outcome studies of school-based programs. Results of the meta-analysis reveal that interactive programs are more effective at preventing substance abuse than knowledge-based or attitude-based programs. The interactive style is consistent with problem-based learning, which is enjoying widespread popularity within educational circles. However, while the interactive style of prevention seem effective, few evaluation studies have investigated the process through which adolescents change their behaviors and perceptions regarding substance use and abuse. In order to address this issue, the authors draw on an on-going interactive intervention study of 300 ninth graders, aged 15 and 16 years, in the Aland Islands, Finland. The evaluation study revealed that the students’ perceptions of the risks involved with substance use could not be altered, nor could their use of alcohol or tobacco during the school year be curbed. However, the students’ expectations regarding alcohol use were changed, which could possibly lead to differences in their patterns of alcohol use. Based on this analysis, the article offers a combination of outcome and process evaluation techniques that may better reveal how school prevention programs can achieve their desired results. References