Most clinical MRSA (methicillin-resistant S. aureus) isolates exhibit low-level β-lactam resistance (oxacillin MIC 2-4 μg/ml) due to the acquisition of a novel penicillin binding protein (PBP2A), encoded by mecA. However, strains can evolve high-level resistance (oxacillin MIC ≥256 μg/ml) by an unknown mechanism. Here we have developed a robust system to explore the basis of the evolution of high-level resistance by inserting mecA into the chromosome of the methicillin-sensitive S. aureus SH1000. Low-level mecA-dependent oxacillin resistance was associated with increased expression of anaerobic respiratory and fermentative genes. High-level resistant derivatives had acquired mutations in either rpoB (RNA polymerase subunit β) or rpoC (RNA polymerase subunit β') and these mutations were shown to be responsible for the observed resistance phenotype. Analysis of rpoB and rpoC mutants revealed decreased growth rates in the absence of antibiotic, and alterations to, transcription elongation. The rpoB and rpoC mutations resulted in decreased expression to parental levels, of anaerobic respiratory and fermentative genes and specific upregulation of 11 genes including mecA. There was however no direct correlation between resistance and the amount of PBP2A. A mutational analysis of the differentially expressed genes revealed that a member of the S. aureus Type VII secretion system is required for high level resistance. Interestingly, the genomes of two of the high level resistant evolved strains also contained missense mutations in this same locus. Finally, the set of genetically matched strains revealed that high level antibiotic resistance does not incur a significant fitness cost during pathogenesis. Our analysis demonstrates the complex interplay between antibiotic resistance mechanisms and core cell physiology, providing new insight into how such important resistance properties evolve.