The completeness of vagotomy following proximal gastric vagotomy or anterior seromyotomy with posterior truncal vagotomy was assessed prospectively in 48 patients using the intraoperative congo red test. Pentagastrin (6 micrograms/kg) was given subcutaneously before the assessment. An endoscope was passed into the stomach and 180 ml congo red solution washed over the gastric mucosa. Continuing acid production was indicated by the appearance of a black colour (pH < 3) 2 min after introduction of the dye. A grading system was adopted where grades I and II showed little black discoloration and grades III and IV showed increasing areas of discoloration indicating that further denervation was required. All 20 patients undergoing anterior seromyotomy with posterior vagotomy were classified as grade I. Fifteen of an initial 23 patients receiving proximal gastric vagotomy were grade III or IV. Following division of either the right gastroepiploic nerve or the posterior vagal trunk, 22 patients improved to grade I (16) or II (six). In the subsequent five proximal vagotomies, modification of the dissection produced grade I results. Anterior seromyotomy with posterior truncal vagotomy gave consistently complete vagotomy. The congo red test highlighted major differences in the adequacy of vagotomy achieved using various dissection techniques during proximal gastric vagotomy. The test is a useful, reproducible and simple intraoperative method for assessing the completeness of denervation.