Evidence that physical inactivity and low fitness confer an increased risk of coronary heart disease (CHD) is convincing. There is a graded relationship with the amount of physical activity (or physical fitness), with some evidence that an asymptote is reached in the mid-range. Epidemiological studies have also shown that physically inactive individuals are at greater risk of developing hypertension or non-insulin-dependent diabetes or of experiencing a stroke, but less is known about the nature of these relationships. The effects of exercise on blood pressure, glucose/insulin dynamics and lipoprotein metabolism may contribute to the lower risk of these diseases in people who exercise regularly. Long-term adaptations to regular exercise may result in improved insulin sensitivity and in higher serum concentrations of high-density lipoprotein cholesterol-mediated in part by improved weight regulation. However, the residual effects of individual exercise bouts may, cumulatively, also be important; these "acute' effects may be enhanced when functional capacity is increased through training. More intensive exercise may carry greater benefits in some respects, but it also carries higher risks, for example of orthopaedic injury or triggering of heart attack. Consequently, public health policies should aim to foster a long-lasting commitment to increased levels of frequent, moderate-intensity activity in as many people as possible.