ABSTRACT: A rap contest methodology for smoking prevention was tested with sixth and seventh grade students in a predominantly minority public school district. Contests were held after initial assemblies in which students heard anti-smoking rap messages from same-age and older peers on audio and video tape. Pretesting and posttesting (N = 268) indicated high preference ratings for most aspects of the intervention. Analyses of variance revealed no differences across races on any of the dependent measures. However, smaller assemblies were more effective than larger ones in enhancing attitudes against smoking and obtaining more positive contest evaluations and predictions about smoking behavior. A rap contest method therefore may be effective against the initiation of smoking by disadvantaged children in sixth and seventh grades because it is highly acceptable and perceived as culturally relevant. This effectiveness may be more domonstrable in single classrooms rather than larger assemblies.