The currently available glucose-containing peritoneal dialysis fluids (PDF), which are all hyperosmolar, are toxic to the cells present in the peritoneal cavity. However, glucose-polymer solutions, being isosmolar, may have improved biocompatibility in this respect. We therefore compared in vitro the effects of PDF containing glucose-polymers with that of glucose solutions on the function of donor granulocytes and monocytes (INN), and on the viability of mesothelial cells. In addition, the function of peritoneal macrophages (PIVIO) of eight patients was studied in a randomized crossover setting following intraperitoneal exposure to glucose-polymer- versus glucose-monomer-containing fluid of comparable ultrafiltration capacity. Donor granulocytes, as well as INN, showed significantly better phagocytosis of both Staphylococcus epidermidis and Escherichia coli after incubation in the glucose-polymer solution as compared with the 3.86% glucose-containing fluid. Their oxidative metabolism, as measured by chemiluminescence, also showed that the glucose-polymer solution was less inhibitory than fluids containing 2.27 or 3.86% glucose. Patient-derived PING showed a significantly better phagocytic capacity for S epidermidis and E coli, a significantly higher killing of E coli, and a significantly higher chemiluminescence response after intraperitoneal exposure to the glucose-polymer solution as compared with the glucose-monomer-based fluid. Increasing the osmolality of the glucose-polymer solution to that of the respective glucose solutions blunted the favorable effect on phagocyte function, suggesting the beneficial effect to be osmolality-mediated. However, no major difference was observed between the glucose-polymer solution and the glucose-based fluid in their effects on mesothelial viability. We conclude that a glucose-polymer fluid has improved biocompatibility compared with glucose-monomer-based solutions. © 1993, National Kidney Foundation, Inc.. All rights reserved.