Methicillin-resistant Staphylococcus aureus (MRSA) is a major cause of hospital-acquired infection, which is associated with increased cost and length of hospital stays and substantial morbidity and mortality. The rapid progression of whole-genome sequencing (WGS) technologies over the past decade has provided highly detailed and discriminatory insights into the epidemiology of MRSA. However, WGS is still not implemented in the routine surveillance of MRSA at a local, regional or national level. This thesis applies WGS to a number of questions and populations to describe its benefits. Standard infection control investigation commonly uses the antimicrobial susceptibility patterns (ASPs, patterns of susceptibility and resistance to commonly used antibiotics) of MRSA as a surrogate for bacterial relatedness. As they are readily available, ASPs are often combined with patient movement to evaluate putative MRSA outbreaks in hospitals. The accuracy of this method was evaluated by comparing linked cases based on MRSA ASPs versus linked cases based on whole-genome relatedness, using data from a year-long prospective observational cohort study of 1,465 MRSA-positive individuals. The sensitivity and specificity of ASP in the presence of a direct ward contact was 44% and 85%, respectively; in the presence of a shared residential post code, the sensitivity and specificity of ASP is 59% and 76%, respectively. This demonstrates that compared to WGS plus epidemiology, ASP and epidemiology does not reliably identify or refute transmission events. A lineage referred to as epidemic (E)-MRSA15 is largely considered to be associated with hospital settings, but an epidemiological and genomic investigation of a MRSA outbreak in a General Practice (GP) identified 15 people who were E-MRSA15 positive, the majority of which shared a link to a leg ulcer/podiatry clinic in the GP surgery. The outbreak had not been detected previously, and was only identified post-hoc from MRSA sequence data, highlighting the importance of MRSA sequencing to detect otherwise cryptic community outbreaks. The utility of WGS for the investigation of regional MRSA epidemiology was explored for two potentially high-risk MRSA lineages (USA300 and ST2371) that are otherwise unmonitored by current surveillance. Screening the genomes of MRSA isolated from 1,465 people identified over a 12-month period demonstrated that 4.2% of cases were positive for MRSA USA300, with multiple introductions and household transmissions identified. Five people were positive for ST2371, all of whom had a direct or indirect link to a substantial outbreak in an intensive care unit at Addenbrooke’s Hospital in 2011, thus confirming the value of WGS in regional epidemiological investigations. The feasibility and utility of incorporating WGS into routine national MRSA surveillance was evaluated through a combined epidemiological and genomic survey of MRSA bacteraemia undertaken in England over a one-year period. This captured 903 reported cases of MRSA bacteraemia, with 425 isolates available for sequencing. Almost two thirds of isolates were assigned to multi-locus sequence type clonal complex (CC) 22. The addition of MRSA genomes from published outbreak investigations showed that the study genomes could provide context for outbreak isolates and supported cluster identification. Potentially high-risk lineages were also detected. These findings support the integration of epidemiological and genomic surveillance for MRSA bacteraemia as a first step towards a comprehensive national surveillance programme.