One hundred and seventy-one patients were treated by vagotomy and drainage for duodenal ulcer. When performed electively, the mortality was low (0.9%). For massive, continuing hemorrhage the mortality was high (22.2%). There was an 11.1% overall incidence of ulcer recurrence. The number of vagal trunks resected did not establish the completeness of vagotomy. Insulin testing was performed postoperatively in 60 patients. Of these, 22 (37%) were positive. A positive insulin test was associated with a high incidence of recurrent ulcer (60%). There were no recurrences in those patients with negative or equivocal Hollander tests. We believe that routine postoperative insulin tests are a valuable prognostic guide.