The Burge and Grassi tests were compared in 22 patients after parietal cell vagotomy. Dissection was continued until the Burge test was negative or until no further nerve fibres could be found. Three Burge-negative patients were Grassi-positive, 2 due to inadequate division of the terminal branches of the nerve of Latarjet at the 'crow's foot' and 1 due to epiploic fibres. One Burge-positive patient was also Grassi-positive but the latter test indicated the position of an undivided oesophageal nerve fibre. It is concluded that the Grassi intraoperative test for completeness of vagotomy is superior to the Burge procedure because it shows how much stomach needs to be vagotomized to denervate the parietal cells and abolish gastric acid secretion and also indicates the position of residual fibres if vagotomy is incomplete.