Background: We hypothesized that regimens with an early switch to oral antibiotics are as effective as prolonged parenteral regimens for staphylococcal osteomyelitis. Methods: We retrospectively reviewed records of adult patients with osteomyelitis caused by Staphylococcus aureus as determined by sterile site cultures, who had at least 6 months of follow-up post-therapy. The population was divided into two treatment groups: (1) an intravenous group (IV) that received ≥ 4 weeks of parenteral therapy, and (2) a switch group that received < 4 weeks of intravenous followed by oral therapy. Results: A total of 72 patients (36 in each group) were identified with groups evenly matched for demographic and clinical characteristics. The overall apparent cure rate was 74%; 69% for the IV group and 78% for the switch group (P = 0.59). Apparent cure rates were similar regardless of duration of intravenous therapy: 83% < 2 weeks, 72% 2-4 weeks, 75% 4-6 weeks and 66% ≥ 6 weeks (P = 0.68). Among the 39 patients who received rifampin-based combinations, those treated simultaneously with vancomycin and rifampin did significantly worse than those who received other rifampin combinations (P < 0.02). Overall, MRSA infections responded poorly compared to MSSA (65% apparently cured versus 83%). However, 11/14 (79%) MRSA patients who received rifampin combinations, other than vancomycin and rifampin simultaneously, were apparently cured. Conclusions: Overall outcomes did not differ significantly between IV and switch groups. Given the markedly lower costs and ease of administration, prolonged oral regimens after initial intravenous therapy may be a preferred regimen for staphylococcal osteomyelitis. © 2006 The British Infection Society.