Peak exercise oxygen consumption (peak VO2), as measured by expired gas analysis, is an accurate, reproducible and reliable method for determining exercise capacity. In this study, a cohort of 468 patients with type 2 diabetes underwent graded exercise testing to measure peak VO2 at baseline; the cohort was followed for five years for the occurrence of cardiovascular disease (CVD) events. Patients who developed CVD events during the five-year follow-up period were found to have significantly lower baseline peak VO2, as compared to those who did not (p = 0.02). Analysis by gender showed that the mean peak VO2 in male patients who developed CVD events was significantly lower than the peak VO2 in those who did not (p < 0.03). Multiple Cox regression analysis also showed low peak VO2 to be an independent factor. In conclusion, patients with type 2 diabetes mellitus with reduced peak VO2 during exercise have a greater tendency to develop future CVD events.