Using random assignment of students to two intervention groups and a comparison school sample, the researchers evaluated a three-group school-based violence prevention program. The three groups were (1) a whole-school intervention, (2) whole-school, cognitive-behavioral and cultural enrichment training, and (3) no violence prevention. The evaluation yielded significant between-group differences by gender from Times 3 to 4. Males showed no significant across-group differences. Females showed a moderate beneficial effect size for perpetration in group 2. Females also exhibited a large beneficial effect for self-reported victimization. Researchers may need to consider gender as a variable for designing inner-city interventions like this. (Contains 2 tables and 3 figures.)