Glucose is used in peritoneal dialysate to produce the gradient for ultrafiltration. The peritoneal membrane's low reflection coefficient for glucose imposes a demand for high transmembrane concentrations, perhaps adding unwanted body burden of glucose. A polymer with a lower permeation rate used as an osmotic agent would circumvent this. We evaluated the mass transfer coefficient (mtc), T1/2disappearance from the peritoneal cavity and ultrafiltration capabilities of a 900 dalton (Mn) starch derived polymer. We compared an 8% (455 mOsm/L) and a 10% (484 mOsm/L) polymer (Pol) solution to available dialysate solutions containing 2.5% (399 mOsm/L) and 4.25% (491 mOsm/L) × glucose (Glc). The dialysate compositions were otherwise similar. Using a randomized complete block design, 5 anephric dogs maintained on chronic peritoneal dialysis were studied. The mtc (ml/min) was greater for the glucose than the polymer solutions (p<0.05): 2.5%-13 and 4.25%-14 vs 8%-5 and 10%-6. The T1/2disappearance (min) was also greater (p<0.05): 2.5% Glc-112 and 4.25% Glc-111 vs 8% Pol-281 and 10% Pol-252. Over a 180 min. period the 2.5% glucose solution generated the least volume of ultrafiltrate (ml, p<0.05):2.5% Glc-113 and 4.25% Glc-589 vs 8% Pol-640; 10% Pol-912. We conclude that the lower permeation rate of the polymer yields ultrafiltration at a lower dialysate osmolality. A polymer solution may be a feasible alternative to glucose.