Two major types of permanent loss of ultrafiltration capacity (UFC) were previously distinguished among patients treated with CAPD: 1) type HDR with high diffusive peritoneal transport rate of small solutes and low osmotic conductance,but with normal fluid absorption rate, and 2) type HAR with high fluid absorption rate, but with normal diffusive peritoneal transport rate of small solutes and normal osmotic conductance. However, the detailed pattern of changes in peritoneal transport parameters in patients developing loss of ultrafiltration capacity is not known. Analysis of solute and fluid transport parameters in the same patient before and after UFC loss. Seven CAPD patients who had undergone repeated dwell studies,which were carried out before and/or after the onset of UFC loss. Dialysis fluids (2 L) with glucose or a mixture of amino acids as osmotic agent at three basic tonicities were applied during 6 hour dwell studies. Fluid and solute transport parameters were previously shown not to be affected by these dialysis solutions (except by hypertonic amino acid-based solution). Intraperitoneal dialysate volume and fluid absorption rate were assessed using radiolabeled human serum albumin (RISA). Osmotic conductance (a(OS))was estimated by a mathematical model as ultrafiltration rate induced by unit osmolality gradient. Diffusive mass transport coefficients, K(BD), for glucose,urea,and creatinine were estimated using the modified Babb-Randerson-Farrell model. Five patients had increased K(BD) for small solutes after the onset of UFC loss,and three of them had decreased a(OS),whereas two patients had normal a(OS). In one of them, a(OS) decreased with time after the onset of UFC loss with concomitant normalization of glucose absorption. In all studies of these five patients the fluid absorption rate was within the normal range. Two other patients had increased fluid absorption rate (about 5 ml/min),and one of them also had increased K(BD) for small solutes,in two consecutive dwell studies in each patient with the second study being carried out at 1 and 7 months respectively after the first one. In all four studies in these two patients, the a(OS) was within the normal range. The sodium dip during dialysis with 3.86% glucose-based solution was lost, not only among most patients with UFC loss related to reduced osmotic conductance, but also in patients with increased K(BD). The occurrence of two major types of UFC loss was confirmed. However, a case of a mixed type of UFC loss with high fluid absorption rate and high K(BD) for small solutes, but normal osmotic conductance, and with normalization of initially high K(BD) for small solutes, linked with decreasing initially normal osmotic conductance,was also found. As a reduced sodium dip with hypertonic glucose solution is not only seen in patients with reduced osmotic conductance, it cannot reliably be used as a single measure of decreased aquaporin function. Permanent ultrafiltration capacity loss may be a dynamic phenomenon with a variety of alterations in peritoneal transport characteristics.