Changes in plasma total cholesterol, triglyceride, and lipoprotein concentrations were assessed in three male endurance cyclists who consumed isoenergetic diets for 28-day periods in which carbohydrate, polyunsaturated fat, or saturated fat contributed about 50% of the daily energy intake. Dietary cholesterol was similar among the diets. Maximal aerobic capacity was maintained at 62 ml O2/(kg X min). Body weights were held within 3% of admission levels. The polyunsaturated fat diet significantly (p less than 0.05) reduced mean fasting plasma total cholesterol in comparison to the saturated fat and carbohydrate diets (160 versus 254 and 243 mg/dl, respectively). Similarly, the polyunsaturated fat diet depressed (p less than 0.05) mean plasma triglycerides relative to the saturated fat and carbohydrate diets (37 versus 62 and 79 mg/dl, respectively). No significant dietary effects were seen on high-density lipoprotein cholesterol. The observed changes in plasma total cholesterol were not significantly different than the values predicted by the Keys' equation, delta CHL = 1.35(2 delta S-delta P) + 1.5 delta Z. We conclude that under controlled conditions in which physical activity is constant l) dietary lipid differences influence fasting serum lipid and lipoprotein concentrations among men with high energy expenditures, and 2) the Keys' equation gives useful predictions of changes in plasma total cholesterol among vigorous men consuming different types and amounts of dietary lipid.