Two hundred sixty-four kindergarten through third grade children from three public schools in San Diego were randomly assigned by classroom to one of three treatments: (1) a role-play based sexual abuse prevention program, (2) a multimedia child abuse prevention program, or (3) a control group receiving fire prevention training. Results indicate that children receiving a role-play based sexual abuse prevention program significantly improved on a touch discrimination task as compared to subjects in a control condition. Although older children knew more than younger children prior to intervention in this sample, no significant grade differences emerged in amount of pretest to posttest change. It is concluded that prevention programs have the potential to make children less suspicious and more realistic about nurturing touch rather than increasing children's fears. 3 tables, 6 notes, and 25 references (Author abstract modified)