The role of peritoneal dialysis (PD) in renal replacement therapy (RRT) remains unclear. There are no controlled trials to provide hard evidence of its efficacy. Comparative studies with haemodialysis from different centres and countries have given conflicting results even when allowing for case mix. Data from the United States on patients starting or receiving treatment in the late 1980s suggested a worse prognosis for older patients, particularly diabetics receiving PD as compared to HD. Analysis of the USRDS data base for patients starting in the early 1990s shows an improvement in outcome but with no difference in overall mortality. The Canadian registry has recently published data showing a better survival with PD than with HD in the first two years of RRT. Morbidity is similar with both therapies, although hospitalization is increased with PD. Unfortunately long-term technique survival is not as good with PD. However, PD has certain medical advantages, particularly the maintenance of residual renal function that contributes to solute and fluid removal. It may also postpone the onset of amyloidosis. Patients transplanted after previous PD have a decreased risk of early acute renal failure and equally good long-term results when compared to those patients who were on HD before transplantation. The quality of life is as good with PD as with center HD, and there are social advantages to PD including an increased chance of employment, more flexible holidays and avoidance of thrice weekly travel to a dialysis center. PD also has logistical advantages and can be utilized by the majority of new patients. We therefore conclude that PD has potential advantages early in the course of RRT, and should therefore be offered as a first option to all suitable new patients. Whether PD has a major or minor role in later years (> 5) remains unclear.