OBJECTIVES: Ceftaroline fosamil, the prodrug of the active metabolite ceftaroline, is a broad-spectrum intravenous cephalosporin with bactericidal activity against common pathogens causing CAP. The objective was to evaluate the efficacy of ceftaroline fosamil compared with other antibiotics recommended for initial empiric treatment of adults admitted to hospital with moderately severe CAP. METHODS: MEDLINE, Medline-In-Process, EMBASE and the Cochrane Controlled Trials Registry were searched to identify published trials in which ceftaroline fosamil, ampicillin/sulbactam, ceftriaxone (1g or 2g q24h), cefuroxime, cefotaxime, co-amoxiclav, ertapenem, levofloxacin and moxifloxacin were used to treat hospitalised patients with CAP. Primary outcomes were clinical success at test-of-cure visit in the modified intention-to-treat (MITT) and clinically evaluable (CE) populations using a NMA with uninformative priors. Clinical success for each antibiotic was reported with 95% credible intervals (CrI95%). A fixed effects model was used. RESULTS: Twelve studies (24 treatment arms) with a total of 4,647 patients with CAP were included in the analysis. Five studies involving 2,888 patients were included in the MITT analysis and 9 studies with 3,182 patients in the CE analysis. The pooled clinical success rates (CrI95%) in the MITT population for each antibiotic were: ceftaroline fosamil 75.5% (71.3% to 79.4%), ceftriaxone 68.2% (65.6% to 70.6%), levofloxacin 74.3% (67.3% to 80.4%), and moxifloxacin 73.8% (67.1% to 79.7%). Clinical success rates (CrI95%) in the CE population were ceftaroline fosamil 85.9% (81.4% to 89.6%), ceftriaxone 79.7% (77.3% to 82.0%), co-amoxiclav 64.6% (46.9% to 78.4%), ertapenem 79.8% (70.3% to 87.0%), levofloxacin 85.5% (77.2% to 91.2%), and moxifloxacin 81.7% (75.2% to 86.9%).  CONCLUSIONS: The results suggest that ceftaroline fosamil may be at least as effective as other antibiotics used in the initial empiric treatment of adults admitted to hospital with CAP. This analysis is limited by the small number of studies that were found, making controlling for heterogeneity challenging.