The etiology and pathogenesis of Kawasaki disease (KD) is largely unknown. Certain demographic factors and laboratory findings are predictive of the development of coronary artery (CA) aneurysms. The objectives of this study were to determine the epidemiology of KD patients in an urban hospital and determine risk factors associated with their development of CA abnormalities. A longitudinal case series of KD patients admitted to Children’s National Medical Center from 1990 to 2002 was examined. Age, sex, ethnic background, duration of fever prior to diagnosis, address, month diagnosed, and CA abnormalities (ectasia or aneurysms) on echocardiography were recorded. Median household income was obtained from the U.S. Census Bureau Web site. The Student t-test, logistic regression analyses, and the Kruskal–Wallis test were used, with significance assumed at p < 0.05. A total of 302 patients were evaluated. CA abnormalties were found in 27 patients (9%), with aneurysms identified in 13 patients (4%). Age was 2.9 ± 2.4 years (range, 2 months to 14 years). A total of 51 patients (16%) were ≤1 year and 35 patients (12%) were ≥5 years. Ethnic distribution was 54% (164) African American, 24% (72) Caucasian, 9% (29) Asian/Pacific Islander, 8% (23) Hispanic, and 5% (14) Middle Eastern. Only 2/164 (1.2%) African Americans developed CA aneurysms. Neighborhood median income of the cohort was $45,400 ± $21,200 ($52,200 ±$25,800 for patients with aneurysms). A total of 28% of cases clustered between December and January. Cases doubled annually in 1999–2001 compared to 1990–1998 (39 vs 19). Multivariate logistic regression found age between 1 and 5 years [p = 0.045; odds ratio, 0.31; 95% confidence interval (CI), 0.10–0.97] and African American race (p = 0.014; odds ratio, 0.15; 95% CI, 0.03–0.68) to be independently protective against CA aneurysms. Duration of fever prior to diagnosis, considered in 210 patients, was different between patients with and without aneurysms (11 ± 5.3 vs 6.5 ± 3.8 days, respectively, p = 0.0007). Multivariate logistic regression found fever longer than 5 days to be the only predictive factor associated with the development of aneurysms and any abnormality. African Americans had a shorter duration of fever than the rest of the cohort (6.03 vs 7.31 days), (p = 0.0087). The epidemiology of KD at our hospital is similar to that at other centers except for the predominance of African Americans with a shorter duration of fever prior to diagnosis and a decreased incidence of CA aneurysms compared to other ethnicities. The protective nature of African American ethnicity against the development of CA aneurysms raises speculation about the role of genetics and its interaction with immunity in the pathogenesis of KD.