The aim of this study was to compare mortality and incidence of end-stage renal disease (ESRD) in patients with type 1 (insulin-dependent) diabetes who had attended a diabetes center with those who had not. The cohort consisted of a total of 1430 patients diagnosed with Type 1 diabetes at 18 years or younger, and between 1965 and 1979. This population-based cohort in Japan was subdivided into two groups: patients who had visited a large diabetes center in Tokyo (n=162) and those who had not (n=1212). Mortality and incidence of ESRD were compared between the two subgroups as of January 1, 1990. Crude mortality was 1.95 per 1000 person-years (95% CI: 0.49-5.06) for those who had visited the center and 6.05 (4.86-7.41) for those who had not. A multivariate Cox proportional hazard model showed that the patients who had visited the center were three times less likely to die (hazard ratio: 0.31, 95% CI: 0.10-0.98) than those who had not. Crude incidence of ESRD was 1.32 (0.22-4.09) and 5.86 (4.65-7.26) for those who had visited the center and for those who had not, respectively. After adjusting for covariates, the patients who had visited the center were five times less likely to develop ESRD (hazard ratio: 0.19, 0.05-0.78) than those who had not. Education and treatment of type 1 diabetes with an integrated management system under specialists and a multidisciplinary team appears to be associated with a better prognosis.