Objective:To assess the joint use of purified protein derivative (PPD) and delayed-type hypersensitivity (DTH) antigens in screening individuals of unknown HIV serostatus for tuberculosis (TB) preventive therapy eligibility.Design:Population-based survey.Methods:A group of migrant farm workers were screened for HIV and skin-tested with PPD, tetanus toxoid (TET),Candida albicans(CAN) and mumps (MUM) antigens by the Mantoux method. Anergy was defined as a ≤2mm reaction to all four antigens. Eligibility for preventive therapy was defined as a reaction of ≥5 mm to PPD among HIV-seropositive individuals, ≥10mm among HIV-seronegatives, or anergy.Results:A total of 253 out of 271 individuals had sufficient data for analysis. Of these, 15 (5%) were HIV-seropositive; 183 (75%), 175 (72%) and 157 (65%) reacted to TET, CAN, and MUM, respectively, and 113 (47%) were eligible for preventive therapy [108 (44%) PPD-positive, five (2%) anergic]. Use of PPD alone was 95% sensitive for detecting preventive therapy eligibility; PPD plus one DTH antigen was more sensitive (99%) but less specific (range, 69–85%); PPD plus two DTH antigens was most specific (CAN + MUM, 84%; TET + MUM, 93%; and TET + CAN, 100%).Conclusions:In this population with 5% HIV seroprevalence, testing for anergy did not significantly increase the detection of preventive therapy eligibility. The use of two DTH antigens is very sensitive and specific. These results support the recommendation of joint PPD and anergy testing for the screening of HIV-seropositive individuals. Our data also suggest, however, that for individuals whose HIV serostatus is unknown, anergy testing should be considered as a screening tool only if the prevalence of anergy is expected to exceed the prevalence of PPD positivity.