Investigation of two outbreaks of Kawasaki syndrome (KS) in the United States in 1979 and in 1980 revealed no evidence of person-to-person transmission or of a common-source exposure among patients. Questionnaire data showed that KS was more likely to occur in children of middle and upper socioeconomic status than in those of lower status (P less than 0.05 and P less than 0.001 for the respective outbreaks) and that patients with KS had a higher incidence of an antecedent, primarily respiratory illness than did controls matched for age, sex, and race (83% of patients in the first outbreak vs. 30% of one control group, P less than 0.01, and vs. 36% of another control group, P less than 0.02; and 56% of patients in the second outbreak vs. 32% of their controls, P less than 0.02). However, laboratory studies did not identify an etiologic agent for either KS or for the antecedent illness that may be a risk factor for KS.