Background. Hypoalbuminaemia is common in peritoneal dialysis (PD) patients and has an associated high mortality. An excess morbidity and mortality has previously been found in patients with high peritoneal transport. A high peritoneal large pore fluid flux (JvL) results in increased peritoneal loss of protein that possibly contributes to patient morbidity. Alternatively, hypoalbuminaemia and high transport status could be just a marker of capillary pathology associated with atherosclerotic comorbidity. Methods. Peritoneal dialys is capacity computer modelling of peritoneal transport, based on Rippe's three-pore model, was performed to measure JvL in 155 incident PD patients 2-4 weeks after PD initiation. Patient clinical and biochemical status was determined -6, -3, -1, 1 and 6 months after PD initiation, and every 6 months thereafter. JvL was redetermined in prevalent patients 2 and 4 years after PD initiation. Results. JvL was 0.106±0.056 ml/min/1.73 m2 (median 0.094, interquartile range 0.068-0.128). It was correlated to age*** (*P<0.05; **P<0.01; ***P<0.001) (20-30 years 0.079±0.04; 70 years 0.121±0.071), but not to gender. No correlation to diabetic or preexisting renal replacement therapy was seen, but patients with atherosclerosis had higher JvL (0.123±0.06 vs 0.100±0.056*) as had patients with other systemic disease (0.121±0.68 vs 0.100±0.051*). JvL was positively correlated to area parameter (r = 0.41***), and negatively correlated to plasma albumin (-0.36***). Patients were divided into three equal groups: group 1, JvL <0.075 ml/min/1.73 m2; group 2, 0.075-0.11; group 3: >0.11. There was no difference between the groups in p-albumin prior to PD. Immediately after PD start, differences between the three groups appeared (1 month p-albumin: (μmol/l) group 1, 548±83; group 2, 533±86; group 3, 497±78**), and persisted for up to 6 years. No significant change in JvL was seen at 2 and 4 years. Patients with significant albuminuria also had hypoalbuminaemia (<1 g/day: 546±81μmol/l; >2 g/day: 503±54μmol/l). Intermittent PD ameliorated the effect of JvL on albumin losses and clearance. Mortality was increased significantly with raised JvL, independently of age (2 year mortality: Group 1, 10%, group 3, 32%*). There was no overall effect on technique survival, but hypoalbuminaemic group 3 patients had a higher failure rate*. Conclusion. JvL is related to hypoalbuminaemia and mortality after PD initiation. A high JvL seems to be a marker of preexisting vascular pathology, and to cause hypoalbuminaemia after PD initiation. It is suggested that peritoneal albumin loss can have an identical pathogenic effect as urinary albumin loss, by causing an iatrogenic 'nephrotic' syndrome.