D-Xylose transport in the human jejunum was studied in vivo using a standard intestinal perfusion technique, and also in vitro in human jejunal brush border membrane vesicles. Initial D-xylose concentrations were linearly related to D-xylose absorption rates, a finding consistent with passive diffusion. Perfusion of D-xylose with varying D-glucose concentrations were aimed at examining D-xylose-D-glucose jejunal cotransport. D-Xylose absorption rates from a 30 mM D-xylose perfusate did not change significantly when 10, 30, or 60 mM glucose were added (-3.0 +/- 0.62 vs -3.34 +/- 0.71, -3.82 +/- 0.81, and -4.56 +/- 0.72 mM/30 cm/hr, respectively; minus indicates net absorption) suggesting an absence of a cotransport system. In brush border membrane vesicles, xylose uptake was partially inhibited by D-glucose and phlorizin. These data suggest that jejunal D-xylose absorption, at concentrations used clinically, is by passive diffusion, which process completely overrides a minor D-glucose cotransport component. The D-xylose tolerance test, therefore, reflects jejunal mucosal surface area and mucosal permeability to D-xylose and not nutrient carbohydrate absorption.