INTRODUCTION: Methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE) are causing serious nosocomial infections. Tigecycline was evaluated in hospitalized patients with MRSA or VRE infection. PATIENTS AND METHODS: A randomized (3:1), double-blind, multicentre, Phase 3 study compared the safety and efficacy of tigecycline with vancomycin or linezolid in hospitalized patients with MRSA or VRE infection, respectively. Patients were treated for 7-28 days and the test-of-cure (TOC) assessment was made 12-37 days after the last dose. The primary efficacy endpoint was the clinical response (cure, failure and indeterminate) in the co-primary, microbiologically evaluable (ME) and microbiologically modified intent-to-treat (m-mITT) populations at the TOC assessment. RESULTS: For MRSA infection, clinical cure rates in the ME population (n = 117) were 81.4% (70 of 86 patients) with tigecycline and 83.9% (26 of 31 patients) with vancomycin. In the m-mITT population (n = 133), clinical cure occurred in 75 of 100 tigecycline-treated patients (75.0%) and in 27 of 33 vancomycin-treated patients (81.8%). In patients with complicated skin and skin structure infections caused by MRSA, cure rates were similar with tigecycline or vancomycin (86.4% versus 86.9% in ME population; and 78.6% versus 87.0% in m-mITT population). In patients with MRSA infection, nausea or vomiting occurred more frequently with tigecycline than with vancomycin (41.0% versus 17.9%); most cases were mild, with only three patients discontinuing treatment. In patients with VRE (total enrollment, 15), 3 of 3 and 3 of 8 patients in the ME and m-mITT populations, respectively, were cured by tigecycline, compared with 2 of 3 patients in the ME and m-mITT populations treated with linezolid. CONCLUSIONS: Tigecycline is safe and effective in hospitalized patients with serious infection caused by MRSA. There were too few cases of VRE to draw any conclusions.