The community-discussion groups in single, 3 1/2 hour sessions addressed knowledge, attitudes, and sexual behaviors associated with HIV infection. Discussion-group efficacy was then evaluated through the AIDS Prevention Test, a presession/postsession survey designed to assess knowledge, attitudes, and commitment to behavior change in the postsession survey in individual categories. The results of the testing revealed significant improvement in knowledge, attitude, and behavior modification after participation in the discussion groups. Pretest knowledge correlated with pretest behavior and posttest knowledge, but pretest behavior only correlated with the posttest behavior. Analysis of changes form pre- to post discussion showed significant correlation of knowledge and attitude changes to behavior change. The Stop AIDS prevention program and its evaluation process are recommended as a national model for community-based AIDS prevention programs. A sample of the AIDS prevention test is appended. 14 references (Author abstract modified)