OBJECTIVE: To determine the use of chest radiographs in the screening of asymptomatic adults infected with the human immunodeficiency virus (HIV). METHODS: A prospective, multicenter study of the pulmonary complications of HIV infection in a community-based cohort of persons with and without HIV infection. The subjects included 1065 HIV-seropositive subjects without the acquired immunodeficiency syndrome at the time of enrollment: 790 homosexual men, 226 injection drug users, and 49 women with heterosexually acquired infection. Frontal and lateral chest radiographs were performed at 3-, 6-, and 12-month intervals, CD4 lymphocyte measurements at 3- and 6-month intervals, tuberculin and mumps skin tests at 12-month intervals, and medical histories and physical examinations at 3- and 6-month intervals. Pulmonary diagnoses that occurred within 2 months following each radiograph were analyzed and correlated with the radiographic results. RESULTS: Evaluable screening chest radiographs (5263) were performed in HIV-seropositive subjects while they were asymptomatic; of these, 5140 (98%) were classified as normal and 123 (2%) as abnormal. A new pulmonary diagnosis was identified within 2 months following a screening radiograph in 55 subjects. Only 11 of these subjects had abnormal radiographs; the sensitivity of the radiograph was 20%. The sensitivity was similarly low at baseline, within each transmission category, and in subjects whose CD4 lymphocyte counts were less than 0.2 x 10(9)/L (200/microL). The types of pulmonary diseases that occurred were similar in the subjects with normal and abnormal screening radiographs. CONCLUSION: Screening chest radiography in asymptomatic HIV-infected adults is unwarranted because the diagnostic yield is low.