A meal stimulates the absorption of water and electrolytes from the proximal jejunal lumen. Neither sham feeding nor gastric distention alters this meal-induced jejunal absorption, implying no role for the cephalic or gastric phases of digestion. This study tested the hypothesis that the small bowel is the origin of the proabsorptive signal for meal-induced jejunal absorption. Twenty-five-centimeter canine proximal jejunal Thiry-Vella fistulas were constructed, and chronic duodenal catheters were placed. Jejunal absorption studies (n = 72) were performed by luminal perfusion of the jejunal segments with an isotonic buffer containing radioactive carbon-labeled polyethylene glycol. Each study consisted of a 1-hour basal period followed by a 3-hour experimental period. Ten groups were studied: control, orally ingested mixed meal, and 600 ml duodenal infusions of either water, saline solution, protein, lipid, carbohydrate, 150 mmol/L mannitol, 300 mmol/L mannitol, or 600 mmol/L mannitol, each delivered at 10 ml/min over 60 minutes. The control, water, and saline solution groups showed no significant changes in integrated 3-hour jejunal absorption above basal. The ingested mixed meal significantly increased water and electrolyte absorption (p less than 0.0001). The isovolumetric, isocaloric duodenal nutrient infusions of protein, lipid, and carbohydrate all significantly increased jejunal water and electrolyte absorption (p less than 0.0001). The poorly absorbed solute mannitol significantly increased absorption (p less than 0.0001) in a dose-dependent fashion. These results indicate that the proabsorptive signal for meal-induced jejunal absorption originates from or distal to the duodenum. This newly defined enteroenteric response occurs independently of nutrient composition and responds to increasing osmolarity of poorly absorbed solutes such as mannitol.