The vagal afferent system is strategically positioned to mediate rapid changes in motility and satiety in response to systemic glucose levels. In the present study we aimed to identify glucose-excited and glucose-inhibited neurons in nodose ganglia and characterize their glucose-sensing properties. Whole-cell patch-clamp recordings in vagal afferent neurons isolated from rat nodose ganglia demonstrated that 31/118 (26%) neurons were depolarized after increasing extracellular glucose from 5 to 15 mm; 19/118 (16%) were hyperpolarized, and 68/118 were non-responsive. A higher incidence of excitatory response to glucose occurred in gastric- than in portal vein-projecting neurons, the latter having a higher incidence of inhibitory response. In glucose-excited neurons, elevated glucose evoked membrane depolarization (11 mV) and an increase in membrane input resistance (361 to 437 M). Current reversed at 99 mV. In glucose-inhibited neurons, membrane hyperpolarization (13 mV) was associated with decreased membrane input resistance (383 to 293 M). Current reversed at 97 mV. Superfusion of tolbutamide, a K(ATP) channel sulfonylurea receptor blocker, elicited identical glucose-excitatory but not glucose-inhibitory responses. Kir6.2 shRNA transfection abolished glucose-excited but not glucose-inhibited responses. Phosphatidylinositol bisphosphate (PIP(2)) depletion using wortmannin increased the fraction of glucose-excited neurons from 26% to 80%. These results show that rat nodose ganglia have glucose-excited and glucose-inhibited neurons, differentially distributed among gastric- and portal vein-projecting nodose neurons. In glucose-excited neurons, glucose metabolism leads to K(ATP) channel closure, triggering membrane depolarization, whereas in glucose-inhibited neurons, the inhibitory effect of elevated glucose is mediated by an ATP-independent K(+) channel. The results also show that PIP(2) can determine the excitability of glucose-excited neurons.