This study developed and tested a gender-specific intervention for preventing substance abuse among adolescent girls. Delivered on CD-ROM by computer, the program was compared with a conventional substance abuse prevention program delivered live in a group setting. Seventh-grade girls in New York City middle schools completed pretests, and, by school, were randomly assigned to receive either gender-specific computer intervention (GSI) or conventional intervention, and were posttested. Analyses of pretest to posttest gain scores showed GSI girls compared to girls receiving conventional intervention to possess a larger repertoire of stress-reduction methods, to report lower approval of cigarettes, alcohol, and drugs, to identify more unhealthy ways to deal with stress, to report lower likelihood of cigarette use or alcohol consumption if asked to do so by best friends, and to hold stronger plans to avoid cigarettes, alcohol, and drugs in the next year. These modest findings lend credence to the promise of gender-specific, computerized interventions for substance abuse prevention among adolescent girls.