The three-pore model of peritoneal fluid transport predicts that once the osmotic gradient has dissipated, fluid reabsorption will be due to a combination of small-pore reabsorption driven by the intravascular oncotic pressure, and an underlying disappearance of fluid from the cavity by lymphatic drainage. Our study measured fluid transport by these pathways in the presence and absence of an osmotic gradient. Paired hypertonic and standard glucose-dwell studies were performed using radio-iodinated serum albumin as an intraperitoneal volume marker and changes in intraperitoneal sodium mass to determine small-pore versus transcellular fluid transport. Disappearance of iodinated albumin was considered to indicate lymphatic drainage. Variability in transcellular ultrafiltration was largely explained by the rate of small-solute transport across the membrane. In the absence of an osmotic gradient, fluid reabsorption occurred via the small-pore pathway, the rate being proportional to the small-solute transport characteristics of the membrane. In most cases, fluid removal from the peritoneal cavity by this pathway was faster than by lymphatic drainage. Our study shows that the three-pore model describes the pathways of peritoneal fluid transport well. In the presence of high solute transport, poor transcellular ultrafiltration was due to loss of the osmotic gradient and an enhanced small-pore reabsorption rate after this gradient dissipated.