Twenty-four-hour gastric pH measurement was used to study duodenogastric reflux. To differentiate between gastric hyposecretion and duodenogastric reflux, we also measured bile acid concentrations in the gastric juice, and regarded pH increases to above 4 as possible episodes of reflux. The procedure was used in 60 patients, divided into the following groups: (1) control group (ten patients); (2) duodenal ulcer (ten patients); (3) type 1 gastric ulcer (five patients); (4) type 3 gastric ulcer (five patients); (5) bilateral truncal vagotomy plus pyloroplasty (ten patients); (6) truncal vagotomy plus Billroth I partial gastrectomy (ten patients); and (7) truncal vagotomy plus Billroth II partial gastrectomy (ten patients). The amount of reflux (areas of pH greater than 4) in the type 1 gastric ulcer and Billroth I and Billroth II groups was significantly greater than that found in the control, duodenal ulcer, type 3 gastric ulcer and truncal vagotomy plus pyloroplasty groups. The mean concentration of total bile acids was also greater in the gastrectomized patients than in the rest of the groups studied. In the type 1 gastric ulcer group the mean bile acid concentration was similar to that of the control group.