Regular aerobic exercise reduces risk of cardiovascular disease far more effectively than any pharmaceutical agent. The precise mechanisms contributing to these health benefits are unknown. Currently, much of our knowledge regarding the molecular regulators of skeletal muscle phenotype remodeling in response to muscle activity is derived from rodent models. Over the past five years large scale gene analysis has emerged as a promising research strategy for studying complex processes in human tissue. This review will principally discuss the application of large scale gene expression profiling to study the molecular responses to longitudinal aerobic exercise training studies in humans. The focus is largely on the Affymetrix technology platform, as this can be most easily compared, in a quantitative manner, across laboratories. Indeed, there are compelling reasons to adopt a common standard to obtain maximum synergy across complex, expensive and invasive human studies. Direct comparisons between array data sets can be made, and these should be considered novel 'experiments', often providing great insight into disease mechanisms. Weaknesses in existing human studies are identified and future objectives are discussed.