The proposed research program will focus on developing and implementing theoretically derived sexual risk-reduction interventions targeting crack cocaine users who are not in drug treatment. The research will be conducted in two phases, a formative elicitation phase and an intervention evaluation phase. Phase 1 will begin with an exploratory investigation of sexual risk-reduction practices and contexts employing guided interview methodology with 45 participants. This will elicit preliminary information about the classes and contents of behavioral beliefs concerning positive and negative outcomes of practicing different sexual risk-reducing behaviors, barriers and facilitators to adoption of these behaviors, normative beliefs concerning the expectations and practices of significant others, social supports for sexual risk reduction, and influencing characteristics of personal social networks and cultural values. Based on this information, specific risk-reducing sexual behaviors will be identified for further study and potential intervention development, and a comprehensive interview instrument focused on these behaviors will be constructed and administered to a stratified sample of 240 target population members. The resulting data will be used to test an integrated theoretical intervention model and will inform the intervention development process by suggesting appropriate strategies for different individuals and behaviors. In Phase 2 of the research, multi-session behavior-change interventions addressing selected risk-reducing sexual practices in two targeted subpopulations and settings will be developed and evaluated in randomized experimental field studies. These interventions will be explicitly based on the proposed theoretical model of behavioral action and will be designed to maximize consistency of delivery and ease of replication. The effects of behavior-specific theory-linked interventions with HIV testing will be compared to the effects of a brief standardized prevention (control) strategy presented in the context of HIV testing. The specific intervention content and activities to be implemented will depend on Phase 1 findings. The first intervention will target male and female crack smokers accessed through street outreach (N = 825), and will build upon a successful intervention for out-of-treatment drug users developed under NIDA's Cooperative Agreement Program. This trial will investigate the benefits of providing activities and content tailored to individuals' readiness to change versus an intervention addressing all levels of personal readiness to change. The second intervention will target crack-using female sex workers (N = 240) and will utilize sex-worker networks to enhance participants' normative perceptions and provide social support for using condoms with both paying and main partners.