In sharp contrast to the considerable worldwide epidemiologic data available on acquired immunodeficiency syndrome incidence and human immunodeficiency virus (HIV) seroprevalence, there is relatively little information about current levels of HIV incidence rates. The authors suggest a novel approach for estimating current HIV incidence rates based on a single cross-sectional survey and on an epidemiologic model. The approach is based on diagnostic tests for HIV p24 antigen to identify individuals in the preantibody or window period (time between exposure to HIV and appearance of detectable HIV antibodies). Individuals in the preantibody period are likely to have been infected very recently because the duration of the preantibody period is relatively short. The authors report data on the duration of p24 antigenemia prior to HIV seroconversion. This duration together with the prevalence of p24 antigenemia obtained from a cross-sectional survey are used in an epidemiologic model to estimate current incidence rates. This approach of estimating incidence rates may be especially useful in developing countries and high-risk populations in which it is difficult to follow cohorts to identify seroconverters, and in the design of vaccine efficacy studies in which current incidence rates are crucial for calculating sample sizes.