Three recent studies using registry data from the United States, in comparing the mortality risks between peritoneal dialysis (PD) and hemodialysis (HD), have consistently found that elderly diabetic women on PD have a higher mortality risk as compared with their counterparts on HD. Though the cause for this observation is not clear, the phenomenon may be unique to the United States. Alternatively, a selection bias impossible to decipher may be at work in these studies, as none of them have data on comorbidity, nutrition, or adequacy of dialysis. Finally, the possibility that elderly diabetic women are, for some reason, more vulnerable to the ill effects of peritoneal dialysis should be considered. We report here a retrospective analysis of 47 diabetic women, above 55 years of age, with end-stage renal disease, who were started on PD and who later died on dialysis. The primary outcome of interest was cause of death. Demographic details about the patients, comorbid conditions, dialysis adequacy, and biochemical parameters at the start of PD were noted. Death in these patients was attributed mainly to vascular causes, and there appeared to be a high prevalence of peripheral vascular disease. Infection was the next major cause of death, being the primary cause in 14 patients. Of these, only 5 patients had peritonitis. On a Cox regression analysis, only patient age and duration of diabetes at onset of dialysis were found to be predictive of vascular death. No factor was found to be predictive of death from infection. It appears that elderly diabetic women on PD die mainly of the long-term complications of diabetes.