Cardiopulmonary indices were used to evaluate the effect of controlled exercise training prescribed on the basis of the heart rate at the ventilatory anaerobic threshold in coronary artery disease patients with and without impaired left ventricular function. Fifty-two patients aged 38-75 years were divided into four groups. The first three groups included patients with a left ventricular ejection fraction of > 45% at rest, as follows: group 1, 10 patients with single-vessel disease; group 2, 12 patients with two-vessel disease; group 3, 10 patients with three-vessel disease. Group 4 comprised 20 patients with left ventricular dysfunction (ejection fraction < 35%). The left ventricular ejection fraction was assessed by multigated acquisition radionuclear study. All patients underwent a cardiopulmonary exercise test before and after the program which lasted 6-9 months. The variables measured were oxygen consumption (VO2), CO<sub>2</sub> output, minute ventilation, O<sub>2</sub> pulse, and ventilatory anaerobic threshold. Significant improvements in maximal VO<sub>2</sub>, maximal CO<sub>2</sub> pulse, and ventilatory anaerobic threshold level were observed in groups 1 2, and 4 (p < O. 1-0.0001), but not in group 3. These findings indicate that the overall cardiac function, as evaluated by cardiopulmonary indices, improves in patients with one- or two-vessel disease with good left ventricular function and in patients with impaired left ventricular function following an exercise training program. Severe coronary disease seems to limit improvement, even in the presence of a good left ventricular function. The results validate the heart rate at the ventilatory anaerobic threshold as the optimal training heart rate in coronary artery disease patients and the cardiopulmonary exercise test as a sensitive tool for evaluating exercise training results.