BACKGROUND The major complication of Kawasaki coronary disease is myocardial infarction caused by thrombus formation inside the aneurysm or by organic obstructive lesion following the regression of aneurysm, while the indications for surgical therapy remain controversial. We have adopted coronary artery bypass grafting (CABG) even in young children for giant coronary aneurysms (more than 8 mm diameter) with or without a stenotic region when myocardial ischemia is detected. We hypothesized that a shorter time-period from diagnosis of acute Kawasaki disease (KD) to CABG would lead to better postoperative results. To elucidate the validity of our strategy, we evaluated preoperative patient characteristics and long-term outcome. METHODS Twenty-one patients (mean age: 12.0 years old) with Kawasaki coronary disease had undergone CABG during the last 12 years. The mean age at the time of acute KD was 2.7 years and the mean time range from diagnosis of acute KD to CABG was 8.1 years. The incidence of preoperative reduced ventricular function was 10 per 21 patients (47.6%). A multivariate logistic regression analysis using patient characteristics showed that the time range from acute KD to CABG was the only predictor for ventricular functional deterioration (p=0.03, odds ratio 1.55. 95%CI: 1.033 approximately 2.325). Based on these results, we divided the patients into two groups of short time range (mean: 3.7 years; group S) and long time range (mean: 13.9 years; group L). RESULTS Preoperative left ventricular functional deterioration was recognized more frequently in group L (9/9, 100%) than in group S (1/12, 8.3%)(p<0.01). Myocardial infarction was documented significantly higher in the group L (6/9, 66.7%) than group S (1/12, 8.3%)(p=0.04). There was no surgical mortality in either group. The arterial grafts demonstrated good potential for growth and graft patency was 96.9%. Moreover, seven of the giant aneurysms proximal to the graft anastomosis showed complete thrombotic occlusion after CABG without development of myocardial infarction. The cardiac events free rate of group L and group S was 66.7% and 100%, respectively, during the postoperative follow up periods of 5.5+/-1.1 years (group L) and 4.7+/-1.1 years (group S). CONCLUSIONS We successfully applied CABG for Kawasaki coronary disease. Based on our experience, a short interval after acute KD appears to be ideal for surgical treatment of Kawasaki coronary disease.