We randomized 389 symptomatic patients with human immunodeficiency virus (HIV) infection to ditiocarb sodium (400 mg/m2 orally for 24 weeks) or a placebo. Patients were well balanced according to Centers for Disease Control (CDC) group, CD4+ cell number, and duration of disease prior to entry. Ten new acquired immunodeficiency syndrome (AIDS)-defining opportunistic infections occurred in the treated patients and 21 in the controls. Reduction of new opportunistic infections in the ditiocarb group was significant in all patients (relative risk [RR], 0.44) and in patients with AIDS (CDC groups IV-C1 and IV-D) (RR, 0.12). The size of the effect of ditiocarb was maintained when data were reanalyzed after exclusion of a patient who progressed to Pneumocystis carinii pneumonia who was not strictly CDC-defined (RR, 0.46), or when considering as new opportunistic infections three events, which were clinically active at entry, but for which the definitive diagnosis was made during study (RR, 0.49). The administration of ditiocarb did not induce any major adverse clinical or biological reactions. We conclude that, in this study, ditiocarb was safe and reduced the incidence of opportunistic infections in patients with symptomatic HIV infection.