The "chronic" effect of exercise on blood pressure has been controversial and the debate has been confused by a large number of studies with inadequate methodology. Recent consistent findings in epidemiological, experimental and longitudinal intervention studies have suggested that a true antihypertensive effect which is independent of confounding effects of sodium intake, weight, etc. is more likely than not. Unlike some other measures of lowering blood pressure such as sodium restriction, alcohol moderation and some drugs, regular exercise is associated with beneficial effects on several risk factors and probably has an independent effect on cardiovascular mortality. The magnitude of the effect in previously sedentary subjects is greater than that of dietary measures which lower blood pressure except for weight reduction in the obese. Long-term effects on blood pressure are supported by evidence of a favourable influence on left ventricular hypertrophy. The mechanisms involved in the antihypertensive effect of exercise are unclear, but sympathetic withdrawal is one factor involved. Present evidence appears sufficient to include regular exercise amongst the useful therapies for hypertension.