Plasma gastrin concentrations and gastric acid output after modified sham feeding were determined in 20 duodenal ulcer patients. Sham feeding produced an acid response corresponding to 40-68% of the maximal acid output after pentagastrin stimulation, with no significant increase of plasma gastrin concentrations. In eight patients proximal gastric vagotomy almost abolished the acid responses to both insulin hypoglycaemia and sham feeding. Sham feeding in the vagotomised patients did not change the gastrin concentrations in plasma. After pretreatment with benzilonium, an anticholinergic with minimal central nervous effects, plasma gastrin concentrations increased after sham feeding. The study confirms that sham feeding is a poor stimulus for gastrin release in duodenal ulcer patients and supports a cholinergic inhibition of gastrin release. Intravenous injection of benzilonium bromide in a dose close to 70 micrograms/kg, and atropine in the low dose of 30 micrograms/kg inhibited the acid response to sham feeding by about 65%. Atropine in a dose of 50 micrograms/kg virtually abolished the acid sham feeding response, possibly owing to ganglionic or central nervous blockade. Vagal activation of the acid secretory glands does not seem to involve a purely cholinergic neurotransmission.