There have been few clinical studies on the association between the vancomycin 24-h area under the concentration-time curve (AUC24) to minimum inhibitory concentration (MIC) ratio and vancomycin treatment outcomes in methicillin-resistant Staphylococcus aureus (MRSA) infections. To examine this association and to establish a suitable cut-off value for AUC24/MIC, a multicentre prospective observational study was conducted in patients with MRSA bacteraemia. Data were collected on all patients aged ≥18 years with MRSA bacteraemia treated with vancomycin for ≥72 h without dialysis. The MIC was determined by broth microdilution (BMD) and Etest. Treatment failure was defined as (i) 30-day mortality, (ii) persistent bacteraemia (≥7 days) and (iii) recurrence (≤30 days after completion of therapy). AUC24 was estimated by a Bayesian approach based on individual vancomycin concentrations. The AUC24/MIC cut-off value for differentiating treatment success and failure was calculated by Classification and Regression Tree (CART) analysis. In total, 117 patients were enrolled, among which vancomycin treatment failure occurred in 38 (32.5%). In univariate analysis, high vancomycin MIC and low trough levels were unrelated to treatment outcomes. In the CART analysis, low vancomycin AUC24/MIC [<392.7 (BMD) and <397.2 (Etest)] was associated with treatment failure. In multivariate analysis, low AUC24/MIC was a risk factor for treatment failure [adjusted odds ratio (aOR)=3.50, 95% confidence interval (CI) 1.39-8.82 by BMD; aOR=5.61, 95% CI 2.07-15.24 by Etest]. AUC24/MIC is associated with vancomycin treatment outcomes in MRSA bacteraemia, and seeking individualised AUC24/MIC ratios above target (>400) may improve treatment outcomes.