The effect of sham feeding on the plasma concentration of gastric inhibitory polypeptide (GIP) was studied in unrestrained rats bearing chronic gastric fistulas and jugular catheters. While no increase of plasma GIP concentration could be detected during sham feeding (fistula open), during normal feeding (fistula closed), plasma GIP concentrations rose rapidly. In contrast to GIP, plasma insulin concentrations showed a rapid and phasic response during sham feeding in the absence of changes of glycemia. In anesthetized rats electrical stimulation of the vagus nerve was without any effect on plasma GIP concentration, while plasma insulin increased rapidly by as much as 150 percent. It is concluded that under the conditions used, the full gastric and/or intestinal phases of food ingestion are necessary to trigger GIP release, and that vagal activation alone is unable to stimulate GIP release in the rat.