In 6 (10 to 12 kg) mongrel female dogs, silver electrodes for recording electrical activity and for pacing of pacesetter potentials (PP) were implanted on the stomach and duodenum and a catheter for intraluminal instillations was inserted into the duodenum. Beginning 2 weeks after operation, electrical recordings were made intermittently from the fasted, conscious dogs with no pacing and during pacing of the PP in the proximal and in the distal duodenum. A suspension of BaSO4 injected into the unpaced duodenum and observed cinefluoroscopically was swept quickly from the duodenum into the jejunum; little or none passed orad into the proximal duodenum, and BaSO4 rarely entered the stomach. Only 1 to 3% of a duodenal infusate of 154 mM NaCl with [14C] polyethylene glycol (2 ml per min) appeared in the stomach after 15 min. The results during proximal duodenal pacing were the same as with no pacing. However, distal duodenal pacing, which reversed the direction of propagation of the duodenal PP's, caused duodenal-gastric reflux of BaSO4 in every dog and forced about 30% of the duodenal infusate into the stomach during fasting and during gastric emptying of 400 ml of 154 mM NaCl; at the same time, the rate of emptying of the gastric instillate was slowed about 25%.