This meta-analysis of 14 studies examines the relationship between abstinence-only programs and the sexual behavior and--attitudes of urban students from middle school to early in college. The results indicate a significant relationship between abstinence-only programs and the outcomes examined. Abstinence-only programs, as a whole, were associated with reduced premarital sexual behavior by nearly 0.4 of a standard deviation unit. For sexual behavior and attitudes combined, abstinence-only programs were associated with improvements of nearly 0.3 of a standard deviation unit. When sophisticated controls were utilized in each of these analyses, the effect sizes were smaller, but were all statistically significant. The importance of these results is discussed.