The broad objectives f this proposal are to understand why inner-city Black female adolescents who are at high risk for HIV infection engage in risky sexual behavior and to identify procedures to deter such behavior. A field experiment will examine whether AIDS knowledge interventions can reduce risk behavior and whether interventions that contain a behavioral skills component designed to modify maladaptive attitudes, increase skill at negotiating difficult sexual situations, and enhance ability and motivation to implement safe sex practices are more effective that knowledge-only interventions. Second, it examines whether interventions with mothers of adolescents can reduce adolescents' risky behavior. Participants will include 600 healthy sexually active Black female adolescent outpatients recruited from the teen clinic of a children's hospital serving a low-income, inner-city community. Adolescents will be randomly assigned to one of three developmentally and culturally appropriate interventions: (a) an AIDS knowledge intervention, (b) an AIDS knowledge plus behavioral skill- building intervention,or (c) a structurally similar control intervention on general health promotion. Orthogonal to this, the adolesecents' mothers are randomly assigned to (a) an AIDS knowledge intervention, (b) the AIDS knowledge intervention plus a behavioral skill-building intervention designed to enhance mothers' ability and motivation to transmit pertinent knowledge and attitudes to their daughters and to communicate with them regarding sexual activity, or (c) a control, general health promotion intervention. The approach draws on Fishbein and Ajzen's theory of reasoned action, Bandura's self-efficacy theory, and applicant's previous AIDS intervention research with Black adolescent populations. Repeated measures ANOVA, planned contrasts, and multiple regression will be used to test experimental effects postintervention and at 3, 6, and 12-month follow-ups. The key outcome variable is self-reported sexual behavior, one not likely to be influenced by self-report biases. Other measures include theoretically relevant variables hypothesized to mediate interventions effects, including AIDS knowledge, attitudes, perceived norms, intentions, and self-efficacy. Analyses will also examine whether interventions effects are systematically different depending on participant characteristics, including age, amount of sexual experience, sexual anxiety, assertiveness, and impulsivity in adolescents and mothers' sexual anxiety. This information would guide the development of interventions tailored to subgroups of the adolescents. Additional analyses will examine whether correlates and precursors of risky behavior suggested by prior research are evident in this population.