During the steady state perfusion of the duodenum, recovery of a duodenal marker (polyethylene glycol 4000) can be used to quantify intraduodenal volumes. We used this technique to measure and time the recovery from the duodenum of a second nonabsorbable marker, phenol red. When liquid test meals containing phenol red were placed in the stomach, the rate of gastric emptying could be quantified from the appearance of the maker in the duodenum. The osmolality of either test meal or duodenal perfusates was then varied so that the osmotic control of gastric emptying could be examined. Gastric emptying was fastest when duodenal contents were isotonic; nonisotonic duodenal contents slowed emptying whether these conditions were achieved by nonisotonic duodenal perfusates or by emptying of a nonisotonic test meal. However, nonisotonic solutions in the stomach and in the jejunum did not slow gastric emptying, so long as duodenal contents were isotonic. Osomoreceptors, which slow gastric emptying, are present in the duodenum but not in the jejunum or stomach.