Aims/hypothesis We investigated the survival rate of Danish diabetic patients with end-stage renal disease (ESRD) between 1990 and 2005 and evaluated possible predictors of survival rate. Materials and methods Data were obtained from the Danish National Register on Dialysis and Transplantation and from the Scandiatransplant database. Survival rates in different patient groups and association with age, sex, calendar time, waiting-list status and renal transplantation were evaluated using a multivariate Cox regression model. Results During the study period 8,421 patients (13% type 1 diabetic, 9% type 2 diabetic and 78% non-diabetic) started renal replacement therapy. The overall survival rate improved by 15% per five calendar years (hazard ratio [HR]=0.85, 95% CI: 0.81-0.88). The percentage of patients within each group who received renal transplantation was: type 1 diabetic: 26%, type 2 diabetic: 5%, non-diabetic: 24%. The survival rate of transplanted patients with diabetes mellitus (types 1 and 2) compared with nondiabetic patients at 1 year was: 95 vs 93%, at 5 years: 80 vs 85% and at 10 years: 52 vs 71%. Among diabetic patients survival rate was better in transplanted than in waiting-list patients (HR =0.21, 95% CI 0.13-0.34), whereas the survival rate in waiting-list patients seemed to be superior to the survival rate among non-transplantation candidates (HR=0.75, 95% CI 0.53-0.1.02, p=0.07). Conclusions/interpretation The survival rate of diabetic patients with ESRD has improved during the last 15 years. Although some selection bias may exist, significantly improved survival rate was observed among transplanted patients compared with dialysis patients on the waiting-list for transplantation. Renal transplantation should therefore be offered to diabetic patients with ESRD whenever possible.