The effect of the presence of food in the intestinal lumen on fluid transport by an intestinal loop isolated from nutrients is debatable and seems to be species dependent. The aim of the present study was to investigate this effect in humans. Fluid and ion transport by a 30-cm-long jejunal loop was measured by the perfusion of a plasmalike electrolyte solution below an occlusive balloon inflated at the angle of Treitz. At the same time, the duodenum was infused at the papilla by saline (control period) or one of the following solutions (test period): protein hydrolysate, starch hydrolysate, lipids, or mixed nutrients. The four solutions (pH 7; 300 mosmol/L; 540 kcal/L) were infused in 6 normal subjects in a randomized order. In 6 further subjects, two other loads of intraduodenal lipids (120 and 1080 kcal/L) were tested according to a similar protocol. Blood samples were taken serially for radioimmunoassays of gastrin, secretin, cholecystokinin, pancreatic polypeptide, gastric inhibitory polypeptide, vasoactive intestinal polypeptide, motilin, and somatostatin. Intraduodenal mixed nutrients, proteins, and lipids significantly reduced water and ion jejunal net absorption or induced a net secretion (without dose-effect relationship for lipids) and stimulated plasma cholecystokinin, pancreatic polypeptide, and gastric inhibitory polypeptide. Intraduodenal lipids also stimulated circulating levels of gastrin and vasoactive intestinal polypeptide. Intraduodenal sugars did not change jejunal fluid and ion transport and significantly increased plasma gastric inhibitory polypeptide. Covariance analysis showed transjejunal fluid movements to be linked with plasma levels of cholecystokinin. We conclude that an intraduodenal mixed meal exerts a secretory effect on a jejunal loop isolated from the nutrients and that this effect is due to the lipid and protein content of the meal; our data are compatible with a mediation of this phenomenon by cholecystokinin.