Male-to-male sexual behavior in India is criminalized in India; many men who have sex with men (MSM) and peer outreach workers are subjected to stigma, harassment, blackmail, sexual assault, violence, and imprisonment (Human Rights Watch, 2002; People's Union for Civil Liberties-Karnataka, 2001). Consequently, MSM in India are at high risk for HIV/AIDS, yet they remain under-reported in epidemiological data (National AIDS Control Organization, 2004) and underserved by HIV prevention. This presentation will identify methodological and ethical challenges in the design and implementation of HIV prevention research with MSM in India. Formative research with vulnerable communities is essential to the planning and implementation of appropriate and acceptable HIV prevention strategies; however, the research itself may meet with challenges due to the same social context that creates a hostile environment in participants' daily lives. Through engagement in a collaborative process, it is possible to design site-specific research methods that meet demands for rigor and representativeness, and which do not compromise the confidentiality or voluntary participation of participants or the safety of participants and research staff. The presentation is based on an anonymous, cross-sectional survey conducted among 200 men recruited from public sex environments (PSEs) in Chennai, India, in order to identify HIV risk behaviors and contexts, HIV/AIDS knowledge, and preferences for preventive interventions. After mapping PSEs, 20 sites were randomly selected; time-space sampling was implemented to recruit a representative sample of high-risk MSM. Research staff offered study cards at PSEs and invited men to participate in an interview the following day. Participants received a small monetary incentive. Face-to-face, structured interviews were conducted by trained staff in private offices at a local non-governmental organization. The presentation will describe a variety of methodological and ethical challenges that were encountered involving sampling, data collection, and protection of participants and research staff, and it will discuss how these challenges were overcome. For example, because of the absence of any identifiable sampling frame the study used site mapping and venue-based time-space sampling to increase the sample's representativeness. Since PSEs were unsafe for both participants and research staff, a data collection protocol was designed so that participants would be invited to interview the next day and rigorous sampling methods would be retained. Since MSM in Chennai faced police blackmail and harassment, even for carrying condoms, the study did not require participants to sign the consent form or to take a copy of the form with them. Instead, copies were maintained at the interview sites. And when an MSM was murdered near two of the recruitment sites, the PI stopped all study activities at those sites. Although local research staff wished to continue with the existing protocol, an alternate protocol was instituted.