A total of 101 patients with proven Staphylococcus aureus infection were included in a double-blind, placebo-controlled study; this study compared oxacillin (12 g/day, intravenously) or vancomycin (2 g/day, intravenously) plus rifampin (1,200 mg/day, orally) with oxacillin or vancomycin plus placebo. We evaluated 65 patients. Of the patients tested, 33 received oxacillin plus rifampin (13 bacteremias), and 32 received oxacillin plus placebo (16 bacteremias). Clinical cure was achieved in 61% of the patients treated with oxacillin plus rifampin and in 56% of the patients treated with oxacillin plus placebo. Improvement was noted in 27 and 25%, respectively, and failure occurred in 9 and 18%, respectively. These differences were not statistically significant. Bacteriological failure occurred in 3 and 28%, respectively (P less than 0.05). None of the failures within the rifampin-treated group was associated with the emergence of a rifampin-resistant mutant. The rates of superinfection were similar in both groups. The geometric means of the serum bactericidal activity after 1, 6, and 11 h were, respectively, 22, 17, and 9 after treatment with oxacillin plus rifampin and 25, 3.4, and 2.3 after treatment with oxacillin plus placebo. It was suggested that the addition of rifampin to oxacillin or vancomycin might only be beneficial to severely ill patients.