Intravenous vancomycin is considered first line therapy for serious methicillin-resistant Staphylococcus aureus (MRSA) infections including bacteremia, central nervous system infection, pneumonia, pleural space infection, bone or joint infection, prosthetic joint infection and deep abscesses. The effectiveness and toxicity of vancomycin depend on its dosing and chosen target. The most recent guidelines suggest targeting area under the curve over 24 hours over minimum inhibitory concentration (AUC/MIC) of 400 to 600. Implementation of AUC/MIC requires Bayesian software that can be variable, costly, complicated and time consuming. Ideally, AUC/MIC dosing would also require susceptibility testing by broth microdilution, which is not commonly done. It is recommended to target AUC of 400 to 600 assuming a MIC of 1ug/mL when MIC by broth microdilution is not known. Targeting a trough level of 10 to 15mg/L may be a reasonable and more practical alternative without compromising effectiveness. We will be conducting a randomized controlled non-inferiority trial to compare intravenous vancomycin dosing strategy targeting a trough level of 10 to 15mg/L versus AUC of 400 to 600 assuming a MIC of 1ug/mL by broth microdilution for serious MRSA infections. The primary outcome will be treatment failure, which is a composite of mortality and microbiologic failure at 90 days. We hypothesize that targeting a trough level of 10 to 15mg/L is non-inferior to targeting a AUC of 400 to 600 in terms of treatment failure. The criterion for non-inferiority is that a two-sided 95% confidence interval for difference in risk of treatment failure will lie within the non-inferiority margin of 10%.