OBJECTIVE: To better understand how community advisory boards (CABs) can be used to improve the quality of HIV prevention trials. DESIGN: Data collected included descriptive and epidemiologic reports, ethnographic observations, and face-to-face semistructured qualitative interviews with 67 CAB and research team members. Interviews were coded for themes related to community-based consultation. SETTING: The study was conducted at 6 sites of the HIV Prevention Trials Network-Los Angeles, California; Birmingham, Alabama; Philadelphia, Pennsylvania; Harare, Zimbabwe; Lima, Peru; and Chiang Mai, Thailand. PARTICIPANTS: Thirty-six CAB members and 31 research team members, identified with the assistance of research staff at each site, were recruited for interviews across the 6 sites. RESULTS: Both "broad community" and "population-specific" models were identified as strategies for CABs to represent potential participants in HIV prevention trials. CABs viewed their role as a bridge between the research team and trial participants. CABs improved prevention clinical trials by assisting in protocol development, recruitment, and retention. In addition, CABs both identified and helped resolve ethical issues in clinical trials. CONCLUSIONS: When given time to develop, CABs appear to be a good strategy for building partnerships between researchers and communities for collaborative research projects. This approach has the potential to build sustainable capacity to identify and address ethical issues in research as well as community needs.