BACKGROUND:There is a deficit of information regarding the factors that influence peritoneal protein excretion (PPE) during PD therapy. In particular, the effects of the modality of PD and other conditions of the dialysis prescription remain unclear. METHOD:This prospective, observational study analysed the effects of prescription characteristics on 24-hour PPE (study variable) in a cohort of patients starting PD. Our statistical analysis included a multi-level mixed model and standardised estimations of peritoneal protein transport during serial four-hour peritoneal equilibrium tests in order to control for disparities in the characteristics of patients managed on different regimens. RESULTS:We evaluated 284 patients, 197 on CAPD and 87 on automated PD (APD), at the start of PD treatment. The two groups differed in terms of clinical characteristics and peritoneal function. Univariate, serial estimates of 24-hour PPE were marginally higher in CAPD patients, and remained essentially stable over time in both groups. Multivariate analyses identified CAPD (B=888.5mg, 95% CI: 327.5/1448.6), total dialysate volume infused per day (B=275.9 mg/Ll; 153.9/397.9) and ultrafiltration (B=0.41 mg/mL; 0.02/0.80) as independent predictors of 24-hour PPE. The model also revealed a minor trend for a lower 24-hour PPE as time on PD increases. CONCLUSIONS:The individual characteristics of peritoneal protein transport are the major determinants of 24-hour PPE. The use of CAPD as the dialysis modality is associated with higher PPE rates than the APD technique, although this difference is counterbalanced by a direct correlation between PPE and the volume of dialysate infused per day. Ultrafiltration and time on dialysis also act as minor independent predictors of PPE during PD therapy.