Abstract The rate of gastric emptying of a 750 ml. 10 percent dextrose meal was studied by a double-dye dilution technique in 23 patients with duodenal ulcer. Fourteen were restudied 3 months after vagotomy with pyloroplasty (V + P) and nine 3 months after vagotomy with antrectomy (V + A). Two points in the emptying were taken to express the results: (1) the volume emptied by 10 minutes after ingestion of the meal as a measurement of initial gastric incontinence or cascading, and (2) the time taken for complete emptying of the stomach as a measurement of active muscular gastric emptying. Statistical comparison of the preoperative and postoperative results by Student's t test showed that both operations significantly shortened the time taken for complete emptying of the stomach (p < 0.025 for V + P and p < 0.05 for V + A). However, only after V + P was there a significant increase above preoperative levels in the volume of early emptying (p < 0.001). After V + A two groups were identifiable: five of the nine patients emptied rapidly, whereas four behaved similarly to the way they did before operation. Rapid early emptying of the meal after either operation correlated with the clinical symptoms of dumping (p < 0.01).