The purpose of this study was to compare high- versus low-intensity exercise training on the change in echocardiographic left ventricular ejection fraction (LVEF) from rest to peak exercise. Sedentary men with coronary artery disease, aged 30 to 70 years, were randomized to dynamic exercise training of either low intensity, 50% of maximal oxygen consumption, n = 89; or high intensity, 85% of maximal oxygen consumption, n = 111. No other interventions were imposed and patients were evaluated at 6 months and 1 year. Both exercise groups significantly increased exercise capacity without adverse events, but the increase was greater (p = 0.02) in the high-intensity exercise group. The mean exercise test rest-peak LVEF in the high-intensity group rose from 6.20% at baseline to 6.54% in 6 months and to 6.73% at 12 months, while the low-intensity group showed no improvement at 6 months and a decrease at 12 months. Multivariate analyses revealed that treatment group (high versus low intensity) significantly contributed to the change in rest-peak LVEF. When the exercise groups were subdivided by initial baseline LVEF < or = 50% versus > 50%, those with the higher LVEF in the high-intensity group showed a greater (p = 0.05) increase in the rest-peak LVEF from baseline to 1 year. Over a 1-year period, exercise capacity improved in both exercise-intensity groups, but more so in the high-intensity group, with no adverse events. The high-intensity group, compared with the low-intensity, showed more improvement in the rest-peak LVEF, especially in those with a higher LVEF at baseline.