Abstract The objective of this study was to assess the comparative effectiveness and safety of antistaphylococcal penicillins (ASPs) and cephalosporins for the definitive treatment of patients with meticillin-susceptible Staphylococcus aureus (MSSA) bacteraemia. PubMed and Scopus electronic databases were searched up to December 2013. All-cause mortality was the primary outcome of interest. A meta-analysis of unadjusted and adjusted data was performed. Seven articles (1643 patients) were included; all but one were retrospective studies, and three of them employed propensity score matching. The studies enrolled primarily adults hospitalised in medical wards for primary or secondary community-acquired, healthcare-associated or nosocomial MSSA bacteraemia. Several ASPs and cephalosporins were compared. Unadjusted 30-day mortality was lower in patients treated with ASPs than in those treated with cephalosporins [risk ratio (RR)=0.62, 95% confidence interval (CI) 0.40–0.98]. Propensity score-adjusted 30-day mortality was not different in patients receiving ASPs or cephalosporins (RR=0.75, 95% CI 0.41–1.39). Substantial heterogeneity and publication bias were found in these analyses. Both unadjusted (RR=0.85, 95% CI 0.54–1.32) and adjusted (RR=1.42, 95% CI 0.22–9.06) 90-day mortality did not differ between patients receiving ASPs or cephalosporins. Limited data regarding adverse events, development of resistance and recurrence were available. In conclusion, the limited available published data derive from retrospective studies and show that there appears to be no statistically significant difference in mortality between ASPs and cephalosporins for the treatment of MSSA bacteraemia.