The administration of intravenous gamma-globulin (IVGG) for Kawasaki disease was investigated throughout Japan in 1993 by obtaining information from pediatric departments in 2652 hospitals that had more than 100 beds. Of 11,221 reported patients, 8958 patients (79.8%) received IVGG treatment. Of all the patients to whom IVGG was administered, the most common total dose was 1000 mg/kg (36.3%) followed by 2000 mg/kg (16.9%) and 1200 mg/kg (16.8%). The treatment was started in 53.8% by day 5 of the illness and in 83.7% by day 7. The proportion of those with cardiac sequelae was higher among patients administered > 2000 mg/kg or in those started on IVGG on day 9 of their illness or later. The possible reasons are (1) those who were more severely affected were treated with high-dose IVGG earlier; or (2) IVGG does not effectively prevent cardiac sequelae. We concluded that there is a risk of unfavorable effects with IVGG regarding cardiac sequelae when the IVGG dose is > 2000 mg/kg or if IVGG is started on day 9 or later. We believe that only a randomized controlled trial, undertaken prospectively, can adequately address the question of the optimal use of IVGG.