SUMMARY Twenty-three consecutive patients with duodenal ulceration complicated by pyloric stenosis who came under the care of one surgeon were treated by highly selective vagotomy (HSV) combined with digital dilatation ofthe stenosis through a gastrotomy. No form of drainage procedure was used. Thus the antral 'mill' and the pyloric sphincter were left intact. Since the stenosis is usually distal to the pylorus rather than truly pyloric such dilatation does not damage the pyloric ring, although it may on occasion lead to perforation of the first part of the duodenum. The subsequent progress of these patients was compared with that of a similar, consecutive series of 23 patients with pyloric stenosis who were treated by truncal vagotomy with a drainage procedure by other surgeons on the same surgical unit. Patients were followed up for between four months and five years. The clinical assessment was carried out in 'blind' fashion at a special gastric follow-up clinic. No evidence of recurrent ulceration was found in either group of patients. Two patients from each group subsequently came to reoperation for the relief of gastric stasis. Twenty-two of the 23 patients (96%) who had undergone HSV plus dilatation eventually achieved a good-to-excellent clinical result (Visick grades 1 + 2), whereas only 17 of the 23 patients (74 %) who had undergone truncal vagotomy with drainage achieved such a result. The main clinical difference between the two groups was that side effects such as diarrhoea and abdominal pain or discomfort were more common after vagotomy with drainage than after HSV. These results bear witness to the remarkable propulsive powers of the gastric antrum after HSV, which were evidently sufficient to overcome any tendency to re-stenosis in more than 90 % of patients. The 9 % incidence of failure due to re-stenosis could perhaps be avoided if a small duodenoplasty were performed instead of simple digital dilatation. The results support the hypothesis that damage to the antral mill and pyloric sphincter can be avoided in the course of operations for 'pyloric' stenosis secondary to duodenal ulceration. Avoidance of the drainage procedure is of benefit to the patient, just as it is in patients who have duodenal ulceration without stenosis.When pyloric stenosis is secondary to duodenal stenosis develops, the stomach compensates for the ulceration the adjective 'pyloric' is not quite accurate, obstruction at its outlet by means of hypertrophy of because in most patients the actual site of narrow-its muscular wall and by forceful peristalsis in the ing is beyond the pylorus in the first part of the antrum. Eventually, as stenosis becomes increasingly duodenum. This distinction is important, because it severe, the stomach gives up the unequal struggle, means that the pyloric sphincter and the antral mill becoming dilated and atonic. remain undamaged by the fibrosis, oedema, and Although truncal vagotomy with a drainage spasm which combine to create the stenosis. As procedure