BACKGROUND: The prevalence of Renal Replacement Therapy (RRT) is rising steadily, worldwide and in Europe. One reason for this is an increasing number of patients starting RRT, but improving survival on RRT may also be contributing. MATERIAL AND METHODS: In an ERA-EDTA Registry study we have examined survival of patients with Standard Primary Renal Disease, or Diabetes, aged 20 to 75 years, who started RRT with haemodialysis (HD) or peritoneal dialysis (PD) between 1975 and 1992. Altogether close to a quarter of a million patients were included in the analysis which included conventional survival analysis of comparable subgroups of the whole cohort as well as Cox regression. RESULTS: After accounting for age, mode of initial treatment, and diagnosis, an improvement in survival of RRT patients was evident. From Cox regression it was calculated the risk for death decreased by about 5% annually during the time period 1975 1992. Patients who started RRT using PD experienced a higher mortality than those starting with HD. According to Cox regression the relative risk ratio for death was 1.25 for the whole period. The difference in survival between patients starting with PD or HD diminished during the observation period (1975-1992). DISCUSSION: The survival prospects of a patient presenting with end stage renal disease were considerably better in the early 1990s compared to the mid 1970s. This is reassuring despite the fact that mortality on RRT remains high. The higher mortality of RRT patients who started with PD is probably an 'historical' observation as the techniques of this treatment modality have improved considerably since the 1980s which was the time period from which came most of the data for the analysis.