Recent epidemiological studies indicate that individuals who maintain a physically active lifestyle are much less likely to develop impaired glucose tolerance and non-insulin-dependent diabetes mellitus (NIDDM). Moreover, it was found that the protective effect of physical activity was strongest for individuals at highest risk of developing NIDDM. Reducing the risk of insulin resistance and NIDDM by regularly performed exercise is also supported by several aging studies. It has been found that older individuals who vigorously train on a regular basis exhibit a greater glucose tolerance and a lower insulin response to a glucose challenge than sedentary individuals of similar age and weight. While the evidence is substantial that aerobic exercise training can reduce the risk of impaired glucose tolerance and NIDDM, the evidence that exercise training is beneficial in the treatment of NIDDM is not particularly strong. Many of the early studies investigating the effects of exercise training on NIDDM could not demonstrate improvements in fasting plasma glucose and insulin levels, or glucose tolerance. The adequacy of the training programmes in many of these studies, however, is questionable. More recent studies using prolonged, vigorous exercise-training protocols have produced more favourable results. There are several important adaptations to exercise training that may be beneficial in the prevention and treatment of insulin resistance, impaired glucose tolerance and NIDDM. An increase in abdominal fat accumulation and loss of muscle mass are highly associated with the development of insulin resistance. Exercise training results in preferential loss of fat from the central regions of the body and should therefore contribute significantly in preventing or alleviating insulin resistance due to its development. Likewise, exercise training can prevent muscle atrophy and stimulate muscle development. Several months of weight training has been found to significantly lower the insulin response to a glucose challenge without affecting glucose tolerance, and to increase the rate of glucose clearance during a euglycaemic clamp. Muscle glucose uptake is equal to the product of the arteriovenous glucose difference and the rate of glucose delivery or muscle blood flow. While it has been known for many years that insulin will accelerate blood glucose extraction by insulin-sensitive peripheral tissues, recent evidence suggests that it can also acutely vasodilate skeletal muscle and increase muscle blood flow in a dose-dependent manner. A reduced ability of insulin to stimulate muscle blood flow is a characteristic of insulin-resistant obese individuals and individuals with NIDDM. Exercise training, however, has been found to help alleviate this problem, and substantially improve the control of insulin over blood glucose. Improvements in insulin resistance and glucose tolerance with exercise training are highly related to an increased skeletal muscle insulin action. This increased insulin action is associated with an increase in the insulin-regulatable glucose transporters, GLUT4, and enzymes responsible for the phosphorylation, storage and oxidation of glucose. Changes in muscle morphology may also be important following training. With exercise training there is an increase in the conversion of fast twitch glycolytic IIb fibres to fast twitch oxidative IIa fibres, as well as an increase in capillary density. IIa fibres have a greater capillary density and are more insulin-sensitive and -responsive than IIb fibres. Evidence has been provided that morphological changes in muscle, particularly the capillary density of the muscle, are associated with changes in fasting insulin levels and glucose tolerance. Furthermore, significant correlations between glucose clearance, muscle capillary density and fibre type have been found in humans during a euglycaemic clamp. Exercise training may also improve control over hepatic glucose production by increasin