The authors assessed the cross-sectional association between intensity of exercise in later life and coronary heart disease risk factors and subclinical disease among 2,274 men and women, 65 years of age and older, who were participants in the Cardiovascular Health Study (CHS) during 1989-1990. Subjects were free of prior clinical cardiovascular disease or impairment of physical function. Exercise intensity was characterized as low, moderate, or high, based on highest intensity exercise reported over the 2 weeks prior to the CHS baseline examination. After adjustment for age, education, and postmenopausal hormone therapy (among women), there was an inverse dose-response relationship of exercise intensity with selected risk factors. By low, moderate, and high exercise intensity, respectively: fasting insulin-men, 15.6 microU/ml, 14.1 microU/ml, and 12.6 microU/ml, p for trend <0.001; women, 14.8 microU/ml, 13.8 microU/ml, and 12.0 microU/ml, p for trend = 0.01; serum fibrinogen-men, 316.2 mg/dl, 315.4 mg/dl, and 300.0 mg/dl, p for trend = 0.01; women, 327.3 mg/dl, 317.0 mg/dl, and 310.7 mg/dl, p for trend = 0.01; lower extremity arterial disease by percent with ankle-arm index <0.9-men, 18.3, 5.5, and 3.7, p for trend = 0.01; women, 10.0, 5.7, and 2.8, p for trend = 0.02; evidence of myocardial injury by cardiac infarction/injury score (CIIS)-men, 8.0, 6.0, 3.9, p for trend <0.001; women, 4.6, 3.9, and 3.6, p for trend = 0.03. Adjustment for smoking, alcohol consumption, and total kilocalories expended in exercise altered the findings only slightly. The authors conclude that intensity of exercise in later life is associated with favorable coronary disease risk factor levels and a reduced prevalence of several markers of subclinical disease.