In 1985 the Division of Mental Health, World Health Organization, Geneva, convened a group of investigators from centers in four countries-Australia, Chile, Norway, Swaziland-to participate in a pilot study on the efficacy of school-based alcohol education. The goal of the educational program was to delay onset and minimize involvement of alcohol use among 13- to 14-year-old adolescents. Twenty-five schools in the four countries, representing middle- and lower-class populations, were randomly assigned to peer-led education, teacher-led education, or a control condition. The educational program was derived from social-psychological theory and etiological research on adolescent alcohol use. The program focused on the social and environmental influences to drink alcohol and skills to resist those influences. It consisted of five lessons over 2 months. Baseline and posttest data measured alcohol use knowledge, attitudes, skills, and friends' drinking patterns. Data were collected immediately prior to and 2 months following the educational program. The data converge on the finding that peer-led education appears to be efficacious in reducing alcohol use across a variety of settings and cultures.