Various small and large molecules have been studied as osmotic agents to replace dextrose in peritoneal dialysis. Macromolecules are attractive because of their slow absorption from intraperitoneal solutions; however, it has been assumed that they are only marginally effective as osmotic agents unless they function as polyelectrolytes at physiological pH. In experimental exchanges conducted in rats we measured volume changes induced by Gelifundol (5.5% oxypolygelatin) and Ringers lactate to which was added either nothing, 4.25% dextrose, or 5% albumin. In the control exchanges using Ringers lactate, intraperitoneal fluid volume remained unchanged for eight hours. The volume changes induced by 4.25% dextrose were complete within two hours and resulted in a two-thirds increase over the amount of fluid administered. In both series in which polyelectrolytes were used volume transport was sustained throughout an eight hour dwell. With 5% albumin the total increase in fluid volume was about 40% of that installed, while Gelifundol caused fluid volume to double. Qualitatively similar results were obtained in transport studies conducted in vitro. Physical studies of the oxypolygelatin solutions indicated that the fixed charges per liter were comparable to those in the albumin solutions. Thus the different volume transport the two proteins induced could not be attributed to Donnan effects. However, since the molecular weight of albumin is triple that of Gelifundol the van't Hoff pressures of the two macromolecules can explain the observed differences in volume transport. These results suggest that neutral macromolecules deserve further study as potential osmotic agents for peritoneal dialysis.