The surgical physiology of the vagus is reviewed with respect to vagotomy in the treatment of duodenal ulcer. All types of vagotomy (truncal, selective gastric, or proximal gastric) produce similar reduction in acid secretion and comparable elevation in serum gastrin. The evidence is mounting that the vagus may have opposing influences on gastrin release: stimulation and inhibition. Division of only the extragastric vagal branches leads to withdrawal of an inhibitory mechanism rendering the denervated stomach more sensitive to the action of gastrin. The loss of this vagally controlled inhibitory mechanism, rather than more meticulous dissection, may explain the higher incidence of more complete vagotomies in selective than in truncal vagotomy. Proximal gastric vagotomy may be the ideal elective operation yet devised for duodenal ulcer. It does, however, cause elevation in serum gastrin and more than 90 per cent of patients after this operation will have positive insulin test in two to four years. This is higher than the positivity seen with truncal vagotomy. Results of controlled trials are needed before this operation becomes fully established.