To evaluate the behavior of exercise parameters in patients with different angiographically defined degrees of revascularization, serial exercise tests were analyzed in 435 patients 1 to 6 years after coronary artery bypass grafting (CABG). All patients had undergone postoperative angiography 2 to 12 months after CABG to determine the degree of revascularization achieved. Revascularization was complete in 182 patients (all significantly stenosed arteries had patent grafts), sufficient in 176 patients (at least the dominant artery supplying the left ventricle had a patent graft) and incomplete in 57 patients (the dominant artery supplying the left ventricle had a closed graft). Twenty patients had all grafts occluded. Exercise tolerance, angina-free exercise tolerance (angina threshold), maximal double product, prevalence of greater than or equal to 0.1 mV exercise-induced S-T segment depression, and the prevalence of the combination of S-T segment depression plus angina pectoris were determined in serial exercise tests (average of 3.0 postoperative exercise tests per patient for a mean follow up of 3.5 years). Patients with complete, sufficient, and incomplete revascularization showed improvement of all exercise parameters for 6, 4, and 1 year after CABG, respectively. Patients with all grafts occluded had improvement of only some exercise parameters. Five years after CABG, exercise tolerance was improved by 24 W (p less than 0.0005) and 21 W (p less than 0.005) in patients with complete and sufficient revascularization, respectively, and not improved in patients with incomplete revascularization or with all grafts occluded. The angiographically determined completeness of revascularization correlates with the extent and the duration of improvement of exercise parameters after CABG.