All available evidence that exercise may protect against coronary heart disease is circumstantial, and direct evidence is difficult to obtain in human beings. Therefore, we studied the effect of moderate conditioning with treadmill exercise on developing coronary-artery disease in monkeys on an atherogenic diet. Physical training was demonstrated by slow heart rates. Serum total cholesterol was the same (approximately 600 mg per deciliter or 15.5 mmol per liter) in exercising and non-exercising monkeys, with significantly higher high-density-lipoprotein (HDL) cholesterol and much lower triglyceride and low-density-lipoprotein (LDL) plus very-low-density-lipoprotein (VLDL) triglyceride in the exercise group. Ischemic electrocardiographic changes, angiographic signs of coronary-artery narrowing, and sudden death were observed only in non-conditioned monkeys, in which post-mortem examination revealed marked coronary atherosclerosis and stenoses. Exercise was associated with substantially reduced overall atherosclerotic involvement, lesion size, and collagen accumulation; it also produced much larger hearts and wider coronary arteries, further reducing the degree of luminal narrowing. Our data suggest that moderate exercise may prevent or retard coronary heart disease in primates.