The effects of endurance exercise training on plasma lipoprotein lipids were determined in 10 men, ages 46 to 62 years, with coronary artery disease (CAD). Patients maintained body weight, health-related behaviors, and stable diets throughout the program. Training was at 50% to 85% of maximal oxygen consumption (VO2 max) for 40 to 60 minutes, 3 to 5 days/week for 29 +/- 7 weeks. Training increased VO2 max (31 +/- 19%, p less than 0.001), reduced plasma cholesterol (C) (-8 +/- 4%, p less than 0.01), low-density lipoprotein-C (LDL-C) (-9 +/- 9%, p less than 0.01), and triglyceride (TG) (-13 +/- 32%, p less than 0.05) concentrations, and increased high-density lipoprotein-C (HDL-C) levels (11 +/- 13%, p less than 0.05) and HDL-C/LDL-C ratios (25 +/- 20%, p less than 0.01). Changes in LDL-C and VO2 max were correlated (r = -0.73, p +/- 0.01), while the changes in LDL-C and HDL-C each correlated inversely with pretraining lipoprotein levels (rLDL-C = -0.77, p less than 0.01; rHDL-C = -0.68, p less than 0.05). Thus potentially "antiatherogenic" benefits of exercise seem to be due to a training effect, since they correlate best with changes in VO2 max and are maximal in patients with initially low VO2 max, high LDL-C, and low HDL-C levels.