The role of Kawasaki disease (KD) as a contributor to early childhood cardiac morbidity in Adelaide was investigated by a review of hospital admission and case-note data from January 1979 to June 1990. There were 57 episodes in 55 patients. The epidemiological data in this South Australian series are similar to that seen in other Australian and New Zealand centres and correlate better with the clinical data from North America than from Japan. The average age of admission was 3.2 years (median 2.7 years) with 38 and 85% of cases being less than 2 and 5 years respectively. The male to female ratio was 1.5. The incidence of KD in the 0-5 year age group was 3.9 cases per 100,000 children. This series represents a minimum number of cases for this period and illustrates an association of aneurysm-risk with prolonged fever; improved defervescence with the combination of intravenous gamma-globulin (IVGG) and aspirin compared with aspirin alone; and a more severe disease process in the very young. The series supports the efficacy of single dose IVGG therapy. Antibiotics were given prior to diagnosis of KD in 79% of patients, often causing diagnostic confusion with possible drug reactions. The pathogenic mechanisms of KD are reviewed and a new hypothesis is proposed that incorporate mechanisms of vessel pathology resulting from release of endothelin and recognized mediators of endothelial damage including tumour necrosis factor-alpha and interleukin-1 beta.