This article presents a conceptual framework for the study of the distribution and determinants of human immunodeficiency virus (HIV) infection in populations, by combining demographic and epidemiological approaches. The proximate-determinants framework has been applied extensively in the study of fertility and child survival in developing countries. Key to the framework is the identification of a set of variables, called "proximate determinants," that can be influenced by changes in contextual variables or by interventions and that have a direct effect on biological mechanisms to influence health outcomes. In HIV research, the biological mechanisms are the components that determine the reproductive rate of infection. The proximate-determinants framework can be used in study design, in the analysis and interpretation of risk factors or intervention studies that include both biological and behavioral data, and in ecological studies.