Many reagents and techniques have been used for delayed-type hypersensitivity (DTH) skin testing in the evaluation of HIV-infected patients, resulting in varied interpretation of the utility of DTH skin testing in this population. We report the development of a simple algorithm for selection of DTH antigens and the clinical relevance of DTH skin testing in HIV disease. Antigens and concentrations for testing were first evaluated in a demographically matched, HIV-negative, immunologically healthy population. The testing scheme was then applied to the HIV population of interest for 5 years at several clinical sites. The antigens and concentrations selected resulted in 100% reactivity to two or more antigens in the HIV-negative cohort. Anergy is thus a distinct immunologic abnormality. Although some correlation (r2 = 0.6) of skin test reactivity and CD4 cell count was found in a cohort of HIV-infected individuals, anergy was found to be independently predictive of the development of symptomatic late-stage disease (Walter Reed Stage 6), AIDS, or death. This stepwise evaluation of skin testing and reagents has led to the modification of the skin testing protocol by defining the minimum number of antigens required and establishing the independent prognostic role of DTH skin testing in the evaluation of HIV-infected patients. The addition of mumps (40 CFU/ml), tetanus (1:10), and Candida (1:10) to the purified protein derivative (PPD) skin test provides the critical controls to evaluate the status of PPD skin test in HIV-infected individuals as well as to provide a useful and prognostic clinical immunology evaluation.