So as to determine the value of brain natriuretic peptide in the plasma as a biochemical marker for myocarditis of Kawasaki disease, we studied 69 patients. The blood samples, electrocardiograms and cross-sectional echocardiograms were obtained before the commencement of treatment and in the convalescent phase. Results: The mean concentration of brain natriuretic peptide in the plasma was 73.2 ± 107.7 (mean ± SD) pg/ml in the acute phase, and 7.9 ± 7.5 pg/ml in the convalescent phase. We checked the electrocardiograms to find abnormal Q waves, elevation or depression of the ST segments, change in the pattern of the QRS complexes, and flattening or inversion of the T wave, all believed to be markers of myocarditis in Kawasaki disease. Those in whom the concentrations were greater than 50 pg/ml in the acute phase showed abnormal electrocardiograms more frequently than did those in whom the values were less than 50 pg/ml (21/29 vs 3/40, p < 0.0001 odds ratio 32.4). Amplitudes of the T wave in standard limb leads were measured both in the acute and convalescent phases, and the differences calculated. We regarded the sum total of these differences as representing “flattening T wave”, and we named this variable as the total suppressed T wave voltage. We examined the correlation between the variable and the levels of brain natriuretic peptide in the plasma during the acute phase, demonstrating a significant correlation (r = 0.500, p < 0.0001). We conclude, therefore, that the concentration of brain natriuretic peptide measured in the plasma can be a useful biochemical marker for the myocarditis of Kawasaki disease. When the titer is over 50 pg/ml, the patient probably has an abnormal electrocardiogram and is most likely to have myocarditis.