In this article, we examine the results from meta-analyses of studies that have focused on the effects of supervised exercise in patients with established type 2 diabetes mellitus. Exercise has been clearly demonstrated to have benefits on blood glucose control (average reduction of glycated hemoglobin, 0.6%) and cardiovascular risk factors. These benefits are observed independently of any change in body mass index and fat mass, and are also seen in older populations. Multiple mechanisms are involved, and the improved insulin-sensitizing effect of exercise training is not restricted to muscle but extends to hepatic and adipose tissue. However, while the benefits of exercise in type 2 diabetes management are undisputable, it is not as easy to draw correlations between clinical benefit and the amount of physical activity included in daily life. Recent studies have shown encouraging results with moderate increases in physical activity, which are feasible for most patients and are sufficient to induce sustained positive changes for 2 years. Thus, the benefits of structured and supervised exercise in patients with type 2 diabetes have been consistently demonstrated. Currently, the primary challenge is to determine how long-term increased physical activity can be durably implemented in a patient's daily life.