We have previously demonstrated that insulin hypoglycemia releases antral gastrin by a pH sensitive mechanism in duodenal ulcer (DU) patients. The effect of vagotomy per se on the hypoglycemic release of gastrin therefore might be obscured by alterations in antral pH. In the present study on 11 DU patients, the gastric acid response to intravenously administered insulin (0.2 units/kg-1) was determined before and after selective vagotomy with pyloroplasty (SV + PP). In another preoperative and postoperative test on each patient, the serum gastrin test, the serum gastrin response (radioimmunoassay) to insulin was determined during gastric perfusion with citrate-phosphate buffer pH 7.0. By adjusting the perfusion rate, the intragastric pH was maintained at 5.0 or higher. SV + PP abolished the acid response to insulin in four and reduced the response by 80% to 95% in another six patients. Gastric buffer perfusion or SV + PP did not alter the basal serum gastrin level. The increase of serum gastrin level after insulin was significantly (P less than 0.01) reduced by SV + PP. Before operation the integrated serum gastrin response to insulin was significant (P less than 0.01). SV + PP reduced the response to one-third. The effect of SV + PP on the hypoglycemic release of gastrin varied among the patients but no relationship was found to the effect on the acid response, nor to the variations of the volume or pH of the perfusate (pH range, 5.0 to 7.5). It is concluded that insulin hypoglycemia releases antral gastrin by a vagal and probably also by a nonvagal mechanism and that both mechanisms are pH sensitive.