This study assessed the effects of 3 theoretically grounded, school-based HIV prevention interventions on inner-city minority high school students' levels of HIV prevention information, motivation, behavioral skills, and behavior. It involved a quasi-experimental controlled trial comparing classroom-based, peer-based, and combined classroom-and peer-based HIV prevention interventions with a standard-ofcare control condition in 4 urban high schools (N ϭ 1,532, primarily 9th-grade students). At 12 months postintervention, the classroom-based intervention resulted in sustained changes in HIV prevention behavior. This article discusses why both of the interventions involving peers were less effective than the classroom-based intervention at the 12-month follow-up and, more generally, suggests a set of possible limiting conditions for the efficacy of peer-based interventions.