HIV infection among women in the U.S. has risen sharply over the past decade. Currently, women are the fastest growing group of HIV infected individuals. HIV and AIDS are spreading fastest among women of color, in particular black women in the U. S. Women become infected predominately through their risky sexual and needle sharing practices related to injection drug use and crack smoking. Interpersonal violence against women, including childhood sexual abuse, rape, and domestic violence are also emerging as major health problems for women in the U.S. Recent research has pointed to a possible link between interpersonal violence and women's drug use and risky HIV practices. The proposed research will examine the link between interpersonal violence and women's drug use and HIV risk and will build upon a recently completed pilot project of 260 women in the Baltimore area. The proposed research is longitudinal in nature. Six hundred black and white women between the ages of 18 and 44 from the Baltimore area will be interviewed annually during the three and a half year study period. Three subsamples of women will be included in the study: an addict sample of 200 injection drugs users and/or crack smokers; a victim sample of 200 women who have experienced at least one of the three types of interpersonal violence; and a stratified area probability sample of 200 women from diverse socio-economic backgrounds. Interviews will be both quantitative and qualitative in nature. A Traumatic Stress/Support model will serve as the theoretical basis for the quantitative portion of the study. Several hypotheses regarding the interrelationships between demographic characteristics, PTSD symptoms, social support, self-esteem, overall life stress, revictimization, and women's drug use and risky sexual and needle sharing practices will be tested using multiple and logistic regression and structural equation techniques. Open-ended life-history interviews, focus groups, and participant observation in and around the social worlds of drug-addicted women will serve as the basis for qualitative study data. Qualitative data will be analyzed using constant comparative and grounded theory techniques.