The aim of this study was to evaluate factors related to the markedly increased risk of dying from diabetic renal disease in Japanese insulin-dependent diabetic patients compared to those in the USA. The study was based on two population-based cohorts consisting of 1374 cases from Japan and 995 cases from Allegheny County, Pennsylvania, USA, who were diagnosed between 1 January 1965 and 31 December 1979. The living status and dialysis experience were determined as of 1 January 1990. The duration-adjusted renal-failure-related mortality rates in the Japanese cohort and the USA cohort were 277.2 and 130.9 per 100,000 person-years, and the duration-adjusted incidence rates of dialysis were 564.9 and 295.6 per 100,000 person-year, respectively. After adjustment for sex, age at onset, calendar year of onset, and duration of diabetes, individuals with insulin-dependent diabetes in the Japanese cohort were still 2.4-fold more likely to receive dialysis compared to those in the USA cohort. Ten of the 36 renal-failure-related deaths in the Japanese cohort had never been treated by dialysis, while all renal-failure-related deaths in the USA cohort had been treated by dialysis. Survival after initiation of dialysis in the Japanese cohort was virtually the same as the USA cohort. These data suggest that a greater frequency of diabetic end-stage renal disease and reduced access to acceptance at dialysis underlie much of the excess of diabetic renal deaths in Japan.