With over 50,000 cases of AIDS and 100,000 cases of HIV infection, the HIV epidemic among injecting drug users (IDUs) in New York City is by far the largest local HIV epidemic among IDUs in the world. Recently, we have observed important positive developments in this epidemic, including: continuing reductions in risk behavior, declining HIV seroprevalence, and low HIV incidence. Whether these positive trends continue will have very important implications for the study of large, high seroprevalence HIV epidemics, which account for the majority of new HIV infections in the U.S. and in industrialized countries. The specific aims for this competing, continuation proposal are to study: 1. Long-term trends in HIV seroprevalence among IDUs in New York City, possible differences in seroprevalence trends in different demographic subgroups, and to monitor estimated HIV incidence. 2. Long-term trends in HIV risk behavior among IDUs in New York City, and characteristics of persons who continue or relapse to high risk behavior. 3. Social network and "mixing" patterns among IDUs, and how these relate to HIV infection and risk reduction. 4. Transitions between non-injecting and injecting drug use, and characteristics of new injectors, including risk behavior, social networks, and HIV, hepatitis B virus (HBV), and hepatitis C virus (HCV) infection. 5. Audio- computer assisted self interviewing as a method for obtaining more complete reporting of HIV risk behaviors. These aims will be accomplished through continuation of cross- sectional surveys of 1000 persons per year entering Beth Israel Medical Center drug treatment programs in New York City. These surveys will include an extensive questionnaire and HIV, HBV, and HCV testing. A random assignment comparison of audio-computer interviewing versus face-to-face interviewing will be embedded in the first 15 months of data collection. The proposed research will provide information of great scientific interest and public health importance regarding the dynamics of high seroprevalence HIV epidemics, the ability of IDUs to sustain risk reduction over long time periods, transitions between non-injecting and injecting drug use, and methods for improving collection of HIV risk behavior data.