BACKGROUND: In patients with left ventricular dysfunction one of the aims of therapy may be to improve functional capacity. This study compared the improvement of functional capacity in response to medical therapy with that caused by revascularization. METHODS: Fifty-two patients with severe left ventricular dysfunction were divided into groups with ischemic cardiomyopathy undergoing revascularization (group A, n = 20) or incremental medical treatment (group B, n = 16) and a control group receiving maximal medical therapy at the start of the study (group C, n = 16). All patients underwent a baseline metabolic exercise test with evaluation of peak oxygen consumption and derived exercise capacity in metabolic equivalents (METS) with standard electrocardiographic and hemodynamic monitoring. Therapy was then optimized in the medical treatment group, whereas the revascularization group underwent coronary bypass grafting. All patients subsequently underwent follow-up metabolic exercise testing. RESULTS: In groups A, B, and C resting left ventricular ejection fractions were comparable (26%+/-5%, 25%+/-7%, and 23%+/-8%, respectively), as were results of initial metabolic exercise tests (4.7+/-0.9 METS, 4.7+/-1.4 METS, and 5.2+/-1.4 METS). At follow-up group A improved exercise capacity from 4.7 to 5.6 METS (p = 0.01). Groups B (4.7 to 5.0 METS) and C (5.2 to 5.6 METS) had no significant improvement. The mean respiratory exchange ratio improved significantly in group A (1.14 to 1.20, p = 0.02), as did left ventricular ejection fraction (26% to 31%, p = 0.02). However, neither parameter changed significantly in groups B or C. CONCLUSIONS: In patients with severe left ventricular dysfunction improvements of exercise capacity are more marked after coronary revascularization than may be obtained after maximization of medical therapy.