Background The epidemiology of Kawasaki disease (KD) shows seasonal variations, although the etiology of KD is unknown. In this study, we compared the clinical epidemiology of KD onset in winter versus that in summer to identify its etiology, that is, infectious agents. Methods Epidemiologic features of KD were compared between two seasons with high incidence (January [winter] and July [summer]) using a dataset of the 22nd nationwide survey in Japan. Data on patients who visited hospital during 2011–2012 in Japan were analyzed after adjusting for age differences. Subgroup analysis was carried out for day of illness at the day of first hospital visit. Results The total number of KD patients reported in the survey was 26 691. The number of patients who visited hospital with KD for the first time in January and July was 2,812 and 2,302, respectively. The proportion of patients in the age group 15 months–3 years was 38.8% in January and 33.5% in July. Mean serum albumin was significantly lower in January than in July (at days 2–5 of illness, P
