According to the hypothesis that most glucose absorption occurs passively across intestinal tight junctions (paracellular absorption), one would predict fairly similar rates of in vivo absorption of L-glucose, the stereoisomer of D-glucose that is absorbed only passively and is not catabolized, and of 3-O-methyl-D-glucose (3OMD-glucose), the D-glucose analogue that is actively and passively transported and not catabolized. In house sparrows Passer domesticus, we applied a pharmacokinetic method to measure simultaneous in vivo absorption of [14C]L-glucose and [3H]3OMD-glucose in a situation in which intestinal glucose transporters were relatively saturated (gavage solution contained 200 mmol l(-1) 3OMD-glucose). Fractional absorptions (F) were not significantly different between [3H]3OMD- and [14C]L-glucose (0.80 vs 0.79), and the apparent rates of absorption did not differ significantly. When we performed the same experiment on other sparrows in a situation in which intestinal glucose transporters were relatively unsaturated (200 mmol l(-1) mannitol replaced 3OMD-glucose in the gavage solution), the apparent rate of absorption was significantly reduced for [14C]l-glucose by 39% and for [3H]3OMD-glucose by 26%. A simulation model showed that a reduction is not predicted if most of the [3H]3OMD-glucose is actively absorbed, because the absorption rate of the tracer should increase when competitive inhibitor (unlabeled 3OMD-glucose) is removed. The similar extent and rates of absorption of [3H]3OMD- and [14C]L-glucose, and the acceleration of their rates of absorption in the presence of luminal 3OMD-glucose, are most consistent with Pappenheimer's hypothesis that the majority of dietary D-glucose is absorbed passively.