Vagal activation produces a gastric acid secretory response by direct nervous stimulation of the parietal cell area and, at least in dogs, by gastrin released mainly from the antrum. In duodenal ulcer (DU) patients antrectomy reduces the acid response to sham feeding slightly more than the maximal acid output in response to pentagastrin, indicating that an antral factor contributes to the acid secretion induced by sham feeding. The marked acid response to sham feeding in antrectomized patients suggests that the direct nervous stimulation of the acid-secreting glands is the predominating stimulus in the vagal activation of acid secretion in man. In the present study vagal activation has been induced by adequate and modified sham feeding and insulin hypoglycemia in DU patients and healthy subjects. The acid response to adequate and modified sham feeding amounted to about 50% of the peak acid output in response to pentagastrin and corresponded to the acid response to an insulin dose of 0.1 U/kg b.w. Modified sham feeding seems to be a simple method of inducing physiological vagal activation of acid secretion. Sham feeding for 15 min increased only insignificantly the plasma concentrations of total gastrin immunoreactivity or heptadecapeptide gastrin. Prolonged sham feeding during intragastric neutralization or sham feeding after proximal gastric vagotomy did not significantly increase the plasma gastrin concentrations. Sham feeding is obviously a poor stimulus for release of gastrin in man. Either release effect of very small amounts of gastrin-17 or release of non-established gastrins may explain the biological effect of an antral factor. Pretreatment with benzilonium, an anticholinergic drug with minimal cerebral actions, increased the gastrin concentration after sham feeding in about half the experiments. This heterogeneous effect supports a non-cholinergic vagal release of gastrin and a cholinergic inhibition of gastrin release but also indicates a complex interaction at the level of the gastrin cells during vagal activation. Evidence for an inhibitory vagogastrone mechanism in DU patients has been found but its effect is weak and transient. Proximal gastric vagotomy abolished the acid responses to both insulin hypoglycemia and sham feeding, in accordance with the view that the direct nervous excitation of the acid-secreting glands is the predominating stimulus in the vagal activation of gastric acid secretion in man. Atropine in low doses or benzilonium inhibited the acid response to sham feeding by only 65%. This finding suggests that the direct vagal excitation of the acid-secreting glands is mediated only partially by cholinergic neurotransmission. Gastric acidification inhibited the gastric acid secretory response to insulin hypoglycemia both in healthy subjects and in DU patients. The inhibitory effect was significantly less in DU patients, however, supporting the concept of a defective inhibition by antral acidification in DU patients.