The permeability characteristics of cat small intestine capillaries were studied using both osmotic transient and lymph-to-plasma protein concentration ratio (L/P) techniques. A vascularly perfused segment of ileum in which superior mesenteric arterial pressure, blood flow, superior mesenteric venous pressure, and lymph flow were monitored was used for both determinations. Intestinal volume was continuously monitored during the osmotic transient experiments. Hyperosmotic solutions of various test solutes were infused at a constant rate directly into the superior mesenteric artery after the establishment of an isovolumetric state. Osmotic conductances were calculated from the initial rate of volume loss and the calculated plasma osmotic pressure change. Reflection coefficients for the various solutes were determined from the osmotic conductance and filtration coefficients within the same preparation. The predicted equivalent pore radius for intestinal capillaries using this approach is 200-350 λ. In another group of cats, the L/P ratio for total plasma proteins was determined prior to and following graded increases in intestinal venous outflow pressure (10-30 mm Hg). At the greatest dilution of lymph proteins, the minimum osmotic reflection coefficient for total plasma proteins (σ(p)) was estimated assuming σ(p)= 1 - (C1/C(p)), and values of σ(p) between 0.87 and 0.92 were acquired. Minimum osmotic reflection coefficients for total plasma proteins were then determined during osmotic transients induced by hypertonic glucose (20-70 mM). Values of σ(p) between 0.56 and 0.74 were acquired during the osmotic transients. The results of these studies suggest that intestinal capillaries are relatively impermeable to endogenous plasma proteins. The discrepancy between the L/P ratio and osmotic transient techniques results, at least in part, from an increased capillary permeability during the osmotic transient.