A prospective randomized trial was conducted to evaluate the effects of exercise-based cardiac rehabilitation after myocardial revascularization surgery (MRS) on work capacity (measured in mets) and left ventricular function as determined from ejection fraction (LVEF). Twenty-eight patients undergoing MRS were randomly assigned to experimental (aerobic exercise, n = 19) or control (muscle relaxation and low-level exercise, n = 9) groups. Patients were studied before surgery (T1) and 2 (T2), 8 (T3), and 24 (T4) weeks after surgery with first-pass radionuclide angiography both while they were at rest and during maximal upright cycle ergometric exercise. Subsets of patients were also studied at T2, T3, and T4 at a standard workload of 75 W, and during maximal exercise 1 year after surgery (T5). Work capacity improved in both groups although significantly more so in the experimental group (3.9, 3.8, 6.0, and 7.3 mets and 3.7, 3.7, 4.9, and 5.7 mets at TI, T2, T3, and T4 in the experimental and control groups, respectively). The differences between groups were significant by T3. Peak exercise LVEF increased significantly in both groups from T1 to T2 then decreased at T3 and remained unchanged through T5. Peak exercise LVEF at T3 to T5 remained significantly above that observed at T 1. LVEF responses were not related to the exercise program. During a standard workload, heart rate decreased, blood pressure increased, and LVEF did not change in either group. After conclusion of the formal protocol (T4), work capacity and LVEF did not change for either group throughout an additional 6 months (T5). We conclude that exercise training significantly enhances both the magnitude and rate of the increase in work capacity after MRS, but that peak exercise LVEF is not influenced by the exercise program. Circulation 69, No. 4, 748-755, 1984. THE GROWTH of rehabilitation programs for patients with cardiovascular disease has been impressive during the 25 years since Hellerstein and Ford first presented "an orderly plan for the rehabilitation of the patient with heart disease. "' Exercise-based rehabilitation programs have a demonstrated beneficial effect on clinical status, as evidenced by improved work capacity,2' 3reduction in incidence of attacks of angina pectoris,4 and indications of a reduction in rates of morbidity/mortality.5 6 Their effect on left ventricular function, which is an important prognosticator of sur-