Eighty-one sedentary but healthy men aged 30-55 participated in a 1 yr randomized study of the effects of exercise on plasma lipoprotein concentrations. Forty-eight were assigned to a running program, while 33 remained as sedentary controls (an approximately 3:2 ratio). After 1 yr the running group had become significantly fitter and leaner than the control group. Lipoprotein concentration changes in the runners (vs. controls) uniformly favored reduced risk of coronary heart disease, but were not significant when all 46 participants with complete data were included. However, the 25 men who averaged at least eight miles (12.9 kilometers) per wk of running increased their plasma high-density-lipoprotein (HDL) cholesterol level by 4.4 mg/dl (p = 0.045) and their HDL2 mass level by 33 mg/dl (p = 0.059), vs. controls. Significant correlations were found for distance run per wk vs. change in plasma HDL-cholesterol (r = 0.48), HDL2 (r = 0.41), and low-density-lipoprotein cholesterol (r = -0.31). Changes in percent body fat and in HDL-cholesterol were correlated (r = -0.47) in runners. There appears to be a threshold at about 8 miles per wk above which a 1-yr running program leads to beneficial lipoprotein changes.