A prospective study of 40 patients undergoing proximal gastric vagotomy confirmed the validity of a new interpretation of the insulin test in the early postoperative period (Maybury et al., 1977). Of the 30 patients falling into the group expected to be free from recurrent ulceration (low risk group), none had developed symptoms suggestive of this condition, whilst of the 10 in the contrasting high risk group, 6 have already developed proved recurrences (mean follow-up period 26 months). A further group of 24 patients was studied by performing a peroperative Burge test for completeness of vagotomy and comparing these results with a postoperative insulin test. This study showed that the Burge test, properly performed, is a reliable test of the adequacy of vagotomy. The use of this test has led to a better understanding of the operative technique required to achieve the objective of an adequate vagotomy.