The inhibitory effect of duodenal exposure to acid and hyperosmolal solutions on pentagastrin-stimulated gastric acid secretion was studied in conscious rats equipped with chronic gastric fistula and duodenal Thiry-Vella loop. The loop was challenged with saline, HCl or hyperosmolal polyethylene glycol. Gastric acid secretion was measured in samples from the gastric fistula. Gut peptide concentrations were measured in duodenal perfusates collected each 30 min, and in plasma samples collected both during stimulated acid secretion alone, and at the end of experiments in combination with luminal challenges of the loops. During pentagastrin-stimulated gastric acid secretion, luminal perfusion of the duodenal loop with acid caused inhibition of acid secretion (P < 0.001) and a prominent release of somatostatin both to the lumen (P < 0.001) and to the circulation (P < 0.05). Also, neurotensin (P < 0.01) and vasoactive intestinal peptide (P < 0.01) were released to the lumen, but not to the circulation. Upon perfusion of the duodenal loop with hyperosmolal polyethylene glycol, acid secretion was inhibited (P < 0.05) and somatostatin alone was released to the luminal side (P < 0.01). In conclusion, duodenal exposure to acid inhibits pentagastrin-stimulated gastric acid secretion and releases SOM to the circulation that may directly inhibit acid secretion. Concomitantly, somatostatin (SOM), neurotensin and vasoactive intestinal peptide are released to the lumen. Duodenal exposure to hyperosmolal polyethylene glycol inhibits acid secretion with a luminal release of SOM only. Thus, luminal acid and hyperosmolal solutions inhibit gastric acid secretion by separate mechanisms. After acid or hyperosmolal challenge, the release of SOM to the circulation indicates gastric acid inhibition in an endocrine manner, while a luminal release of gut peptides indicates a local peptide overflow that might be of importance via paracrine regulatory mechanisms in the intact animal.