Currently, screening tests for HIV infection are designed to detect serum antibodies produced by the immune system in response to protein components of the virus. The enzyme-linked immunosorbent assay (ELISHA) is a test for HIV antibodies developed for large-scale screening of donated blood. The most commonly used confirmatory test is the Western blot assay. The sensitivity and specificity of a test are a function of the cutpoint for declaring the result positive or negative. The number of erroneous test results depends on the population tested, and this paper focuses on three hypothetical populations with low, high, and medium probability of HIV infection to illustrate the importance of the prevalence of disease in the population screened for determining test performance. When assessing the desirability of testing for HIV antibody, the paper recommends that decisionmakers consider the likelihood of correct and erroneous results and the consequences of labeling individuals as seropositive or negative in the use of test results. The burden of proof for establishing an HIV screening program should rest with those proposing it. 6 tables and 43 references.