Nocturnal pH, acid output and volume of gastric secretion have been measured in a group of patients who were referred for surgery because of a poor clinical response to cimetidine. Patients were studied after no treatment, cimetidine 1 g/day and proximal gastric vagotomy. Although pH and acid output were controlled better with cimetidine than no treatment this was not true for volume of secretion. Vagotomy was significantly better than cimetidine in controlling pH, acid output and volume of gastric secretion. The results suggest that cimetidine non-responders should have a good result from proximal gastric vagotomy, making more radical forms of gastric surgery unnecessary.