The long-term outcome of myocardial revascularization by coronary artery bypass grafting in patients with severe coronary obstruction caused by Kawasaki disease is largely unknown. A multicenter follow-up study was performed in 1991. A total of 168 patients with Kawasaki disease (127 male [75.6%] and 41 female patients [24.4%]) who had undergone coronary bypass grafting were enrolled. Obstructive coronary artery disease affected the left main trunk in 11.8%, the right coronary artery in 77.6%, the left anterior descending in 87.6%, and the left circumflex in 25.9%. Old myocardial infarction was noted in 46.0% of the patients. Fifty-four patients (32%, 12.4 +/- 9.8 years) underwent bypass grafting with saphenous vein grafts alone. The remaining 114 patients (68%, 9.8 +/- 7.1 years) received at least one internal thoracic artery graft to the left anterior descending coronary artery. Gastroepiploic artery grafts were used in 12 patients. There were no significant differences between the saphenous vein and internal thoracic artery groups in the mean age at operation (12.4 versus 9.8 years), female ratio (22% versus 25%), the number of patients over 20 years of age (9.3% versus 9.6%), previous history of infarction (51.9% versus 41.2%), impaired left ventricular function (ejection fraction < 0.5) (13.0 versus 11.4%), left main trunk disease (11.1% versus 10.5%), the number of vessels involved (2.2 +/- 0.8 versus 2.0 +/- 0.6 per patient), or the mean number of grafts used (1.7 +/- 0.7 versus 1.7 +/- 0.7 per patient). The operative death rate was also the same in the two groups (1.9% versus 0%), but the late cardiac death rate was significantly higher in the saphenous vein graft group (13.0%) than in the internal thoracic artery group (0.9%) (p < 0.003). Actuarial analysis showed a significantly higher survival in the internal thoracic artery group (98.7% +/- 1.2% versus 81.6% +/- 7.0%, p < 0.05) at 90 months after the operation. Late death was strongly related to the absence of an internal thoracic artery graft (p < 0.003) and to the age at the time of operation (p < 0.05). The actuarial patency rate was significantly higher for arterial grafts (77.1% +/- 1.1%, n = 151) than for vein grafts (46.2% +/- 6.3%, n = 126) 85 months after the operation (p < 0.003). Arterial grafts were used for the non-left anterior descending coronary arteries in only 41 of 155 grafts (26.5%); in contrast, vein grafts were used in 85 of 133 grafts (63.9%) (p < 0.005 to 0.001).(ABSTRACT TRUNCATED AT 400 WORDS)