High Body Mass Index (BMI) has been associated with improved survival of End-Stage Renal Disease (ESRD) patients on chronic hemodialysis (HD); however, studies on the relationship of BMI with survival in Peritoneal Dialysis (PD) patients have yielded conflicting results. The purpose of this study was to evaluate the impact of BMI on survival of Canadian ESRD patients on PD, correcting for their age, sex, race, diabetes mellitus, and arterial hypertension. In an intent to treat study, we reviewed data of the Canadian Organ Replacement Register (CORR), of incident patients, starting PD between 1994 and 1998 and followed up from their initial PD treatment to the end of 2003. Patients were censored at loss to follow up, transplantation, and the end of the observation period. Cox regression (multivariate) analysis was performed and adjustments were made for age, gender, race, primary renal disease and BMI. During these years, 4054 patients commenced PD, 1742 (43%) of them were females and 1471 (36.3%) were diabetics. The majority were Caucasians (n=3058, 75.4%); 120 (3%) belonged to the First Nations, 137 (3.4%) were black, and the rest (739 pts-18.2%) belonged to various other ethnicities. Based on quartiles of the BMI distribution, 1130 patients (28%) had a BMI < 18.5 kg/m(2) stop; 1163 (28.7%), 18.5-24.9 kg/m(2) stop; 1214 (30%), 25-29.9 kg/m(2) stop; 547 (13.5%) > 30 kg/m(2)stop. Intent to treat Cox regression analysis showed that being underweight was a strong risk factor for death. Specifically, a BMI less than 18.5 was associated with a death hazard ratio (HR) 1.3, (CI: 1.1-1.6). On the contrary, BMI > 30 was not associated with worse survival than those with normal BMI (HR = 1.009, CI = 0.89-1.14). High-BMI patients should not be discouraged from PD just because of their size.