Quantitative analysis can reveal the consistency of intervention effects across studies, as well as the variation of effects according to study-level characteristics. After consulting with project experts in methods and content, and reviewing the literatures on research synthesis and on HIV prevention, we developed a systematic protocol of analytical methods for synthesis of behavioral and biologic outcome data from HIV intervention studies. This protocol included procedures for identifying eligible studies; defining, characterizing, and prioritizing outcomes; abstracting and calculating estimates of effect; adjusting for baseline distributions and intraclass correlation; transforming estimates to a common metric; summarizing effects; examining differences in effectiveness among groups of studies; and translating these results into terms useful to HIV prevention practitioners and researchers. We applied these procedures to transform outcome data reported in many different statistical formats into odds ratios that could be combined and compared across studies. We analyzed data on behaviors related to sexual risk for HIV infection (unprotected sex, condom use, and number of partners) as well as data on biologic outcomes (incidence of HIV and other sexually transmitted infections). This framework may be useful for meta-analyses of prevention research in other fields, particularly when primary research features diverse outcome measures and methods of analysis.