Methicillin-resistant Staphylococcus aureus (MRSA) infections are responsible for longer hospital stays, increased hospital costs, and poorer outcomes compared to methicillin-sensitive S. aureus (MSSA) infections. We aimed to describe the epidemiology of S. aureus bacteraemia (SAB) and to determine factors associated with MRSA infection in South Africa.Cases of SAB were reported from September 2012 to September 2013 from three sentinel sites. A case was defined as the isolation of S. aureus from a blood culture during a 21-day period. Detailed clinical information was collected. Multivariable logistic regression was done to determine factors associated with MRSA infection and mortality.There were 442 cases of SAB reported; antimicrobial susceptibility testing was performed on 240 isolates (54%). Thirty-six percent (86/240) of cases had an MRSA infection. A longer hospital stay before positive specimen collection (odds ratio (OR) 1.08, 95% confidence interval (CI) 1.02–1.13, p=0.004), hospitalization in the last year (OR 15.7, 95% CI 2.5–99.5, p=0.003), HIV infection (OR 4.9, 95% CI 1.05–22.90, p=0.044), and antibiotic use in the previous 2 months (OR 0.1, 95% CI 0.01–0.68, p=0.022) were independent predictors of MRSA. Older age, and in particular age 25–44 years (OR 22.2, 95% CI 2.7–185.5, p=0.004, compared to those aged<5 years), was the only independent predictor of mortality amongst cases with SAB. MRSA isolates were non-susceptible to more antimicrobial agents compared to MSSA isolates.HIV infection was an independent risk factor for MRSA infection. The selection of appropriate empirical antimicrobial treatment is essential in patients with MRSA infections because of non-susceptibility to many other antimicrobial classes.