The effects of pyloroplasty and vagal denervation of the distal stomach on gastric emptying of 10% dextrose have been compared. In a randomized trial, 38 male patients having elective surgery were treated by proximal gastric vagotomy (PGV, N = 10), proximal gastric vagotomy and pyloroplasty (PGV + P, N = 9), total gastric vagotomy (TGV, N = 9), and total gastric vagotomy and pyloroplasty (TGV + P, N = 10). Gastric emptying was recorded using an external scanning technique before and after operation. Compared with before operation, all four procedures resulted in significantly faster emptying during ingestion of the meal (the filling phase). This is attributable to failure of receptive relaxation of the proximal stomach. The emptying phase (after ingestion of the meal) was unchanged by PGV, but was significantly altered by PGV + P, TGV, and TGV + P (Principal Component Analysis). Before operation and after PGV, emptying fitted a monoexponential curve. In contrast, after PGV + P, TGV, and TGV + P, the emptying phase fitted a double exponential curve. Emptying was precipitate initially, but slowed later, leaving a large residual volume. We conclude that, when the proximal stomach is denervated, the intact, innervated distal stomach can restore the emptying of hypertonic dextrose to normal; however, when the distal stomach is denervated or pyloroplasty has been performed emptying remains uncontrolled. Since pyloroplasty and vagal denervation have the same effect, it is concluded that the mechanism retarding the emptying of hypertonic dextrose is an active mechanism that resides in the terminal antrum or pylorus.