Various risk factors were evaluated to explain a significantly greater incidence of coronary heart disease in men of Japanese ancestry resident in Hawaii compared with men resident in Japan. The independent predictors of incidence of coronary heart disease in both Japan and Hawaii were systolic blood pressure, serum cholesterol, relative weight and age. These factors appeared to influence incidence similarly in both areas because in each case the correlation coefficients for Japan and Hawaii did not differ significantly. The hypothesis that the greater incidence in Hawaii could be attributed to differences in levels of these risk factors was tested with the Walker-Duncan method. The four variable multiple logistic function describing the probability of coronary heart disease in Japan was applied to the cohort characteristics observed in Hawaii. The estimated incidence thus obtained was not significantly different from that actually observed in the men resident in Hawaii. Therefore the increased coronary risk profile in Hawaii compared with Japan can account for the greater incidence of coronary heart disease in the former. Current cigarette smoking was significantly related to the risk of coronary heart disease in Hawaii but not in Japan. This difference requires further investigation.