This study will test behavioral programs designed to reduce adolescents' "unsafe" sexual behavior. Behavioral rehearsal procedures that emphasize social skills and decision-making will be compared to didactic sex education and a "no-training" control. We believe the social skills training will be more effective for teaching skills to resist peer pressure for "unsafe" sexual behavior and, thereby, reduce the consequent risks of AIDS and other sexually transmitted diseases (STDs). This analysis also will assess the extent to which trained youth influence their friends knowledge and social skills, as an estimate of the "diffusion" of effects on peer pressure/support of responsible sexual development. A Public Health model will be demonstrated by training youth recruited from community medical clinics. A randomized clinical trail (stratified for Hispanic, Anglo, and gender) design will be used. Three hundred youth from 14-16 yrs. and two friends (600 additional youth, nested within the primary sample) will be divided into three groups: one experimental, one usual-training control and one no-training control. Youth assigned to the experimental group will receive 8 weeks of training. A professional instructor and a specially trained peer assistant will use behavioral rehearsal techniques, including video models, role- playing and corrective feedback, to establish target social and decision-making skills. Youth will be taught how to say "no" (abstain from sex), how to request a partner to use a condom, how to select safe sexual partners, etc. The immediate and long term consequences of their social/sexual behavior will be reviewed to establish establish decision-making skills. Usual training controls will be taught only the biological processes of reproduction and the natural history of STDs and explained the most important risk factors involved; no-training controls will receive only the scheduled measures. They will not be taught specific social skills, nor will behavioral rehearsal be employed. The differential effects of training will be measured for the primary sample of 300 youth and 600 friends. Measures of knowledge, affective reactions and social skills will be obtained immediately prior to instruction, post instruction, at 3, 6, 12 and 18 months. Knowledge, affective reactions and sexual experience (e.g. intercourse) will be measured in private one-to-one interviews. Social skills will be measured by direct observation of role-played tests. These techniques may provide youth with the skills to avoid AIDS infection, and other STDS.