Recent studies suggest that HIV epidemics are diffusing out from injecting drug using communities into the general population through heterosexual transmission, the extent & speed of diffusion remains unknown. The objectives of this study are: (1) to assess the likely extent of such diffusion & (2) to provide an objective & logical framework for making rational comparisons of different ways of deploying behavioral & medical intervention strategies. To achieve these objectives I have five specific aims: . to understand the dynamics of heterosexual & injecting drug use transmission within injecting drug using communities. . to quantify the epidemiological impact of the sexual mixing between injecting drug users (IDU) & their non-injecting sex partners. . to understand the effect that past & current changes in injecting & heterosexual behavior have had (& will have) on seroprevalence levels & AIDS incidence rates. . to assess the epidemiological consequences of a variety of possible behavioral intervention strategies that change injecting &/or sexual behaviors in IDU & their non-injecting sex partners. . to assess the epidemiological consequences of a variety of possible medical intervention strategies in IDU & their non-injecting sex partners. I will assess the effects of vaccines that prevent infection & the effects of therapeutic treatments that slow disease progression. I plan to adopt two complementary approaches to achieve my specific aims: to develop & to apply data analytic techniques to specific data sets, & to formulate & analyze HIV transmission models. I am collaborating with Dr. Schoenbaum in New York City, Dr. Moss in San Francisco & Dr. Coutinho in Amsterdam, The Netherlands. All three of these investigators are in charge of cohorts of injecting drug users, all three cohorts have been followed for a number of years & were designed as longitudinal studies of HIV transmission. I will collaborate closely with these investigators & use their data sets to estimate parameter values for the transmission models & to identify mechanisms of risk behavior change. The HIV epidemics due to injecting drug use & heterosexual transmission have been significantly different in NYC, San Francisco & Amsterdam. I am proposing to use data from the three cities to explain this geographic variation. I am also collaborating with Dr. Nickolas Jewell in my development of data analytic techniques.