Methicillin-resistant Staphylococcus aureus is a leading cause of hospital infections world-wide. Consecutive S. aureus wound isolates (n=561) were collected from PHW Microbiology ABM Laboratory, Swansea (PHW-ABM); 137 (24.4%) were mecA-positive; 424 (75.6%) were mecA-negative using real-time PCR. Audit revealed that 15 (10.9%) mecA-positive strains were not reported as MRSA. Genotyping was performed using pulsed-field gel electrophoresis (PFGE), spa typing and SCCmec typing. MRSA predominantly belonged to EMRSA-15 (89.1% >) and EMRSA-16 (5.8%) clones. All S. aureus strains were included in an evaluation of three SCCmec-orfX PCR assays. The assays had high diagnostic sensitivity (>95%) and specificity (> 94% >) but false negative and false positive results were obtained. A deletion at the SCCmec-orfX right junction was proposed as the probable cause of false negative results. SCCmec-associated loci ccrAB1, ccrAB4, ccrC, and dcs were detected in four false positive MSSA, respectively. MALDI Biotyper mass spectrometry was evaluated for identification of S. aureus. Nineteen (3.4%) of the PHW-ABM wound isolates were Panton Valentine-Leukocidin (PVL)-positive S. aureus. The molecular epidemiology of these and PVL-positive S. aureus (n=61) from Specialist Antimicrobial Chemotherapy Unit, Cardiff (SACU) was investigated using mecA and arginine catabolic mobile element PCRs, PFGE, spa and SCCmec typing. The PHW-ABM strains were predominantly MSSA belonging to the CC159 (n=5; 26.3%), CC275 (n=4; 21.1%) and CC005 (n=2; 10.5%) spa-BURP clusters, affiliated to the ST121, ST30 and ST22 lineages. Within the SACU cohort the USA300 clone (n=16; 26.2%) was predominant, other genotypes included: t044- MRSA-IVc (n=5; 8.2%); t002-MRSA-IVc (n=3; 4.9%) and t127-MRSA-IVa (n=2; 3.3%), affiliated to the European (ST80), USA800 (ST5) and USA400 (ST1) clones. Susceptibility testing demonstrated statistically significant differences between the SACU and PHW-ABM cohorts for oxacillin 57%/5% >, gentamicin 2%/16%, and tetracycline 10% >/42% resistance (p < 0.05). These observed differences highlight the importance of including unselected strains in addition to referred reference laboratory isolates in epidemiological investigations.