Six gastrointestinal hormones were measured in the plasma of six healthy controls and long-term changes were evaluated in six patients 2-20 years after upper gastrointestinal surgery. In a metabolic unit study we determined fasting hormonal levels, the time to peak hormonal response, and a 135-minute hormonal response to the meal. Test meals were isocaloric, 500 kcal, and isonitrogenous, consisting either of natural breakfast components or of complete liquid diets with intact protein (Ensure) or hydrolyzed protein (Vital). Postsurgical subjects were in good health and had no postcibal complaints. Nevertheless, their hemoglobin and serum albumin were significantly lower than in controls. Postsurgical subjects had higher fasting gastrin (121.3 +/−11.6 vs 65.4 +/−6.6 pg/ml, P less than .01) and motilin (148.7 +/−32.9 vs 70.4 +/−13.1 pg/ml, P less than .05) than controls. In postsurgical patients the peak gastrin and pancreatic glucagon responses to meals were obtained in significantly shorter time. Their total response to motilin and secretin to meals was significantly lower than in controls. Fasting glucose and the meal-induced responses of insulin and vasoactive intestinal polypeptide were not different from controls. The nature of dietary protein did not significantly affect hormonal responses to feeding. We conclude that gastrointestinal hormonal changes persist many years after surgery. These changes are probably related to faster transit of meals with a generally weaker total hormonal response to feeding. Although these differences from normal may be nutritionally well compensated, they may become important in periods of metabolic stress.