Opioids, both endogenous and exogenous, have been shown to exert a modulating influence on many gastrointestinal functions. Animal studies suggest an inhibitory role for opioids in the control of pancreatic secretion, which may be exerted through their inhibitory effect on acetylcholine release from intramural neurons. We studied pancreatic secretion in six healthy male volunteers (mean age 42 years) by means of duodenal aspiration using PEG-4000 as a recovery marker. Pancreatic secretion was stimulated with duodenal perfusion of mannitol (510 mosmol/kg) and an equimolar solution (0.3 M/liter) ofl-phenylalanine andl-tryptophan and also with an intravenous bolus of secretin and cholecystokinin (each 1 CU/kg). Each subject underwent six studies on separate days with intravenous infusion of saline, naloxone 40 ?g/kg/hr, morphine sulfate 40 ?g/kg/hr, morphine and naloxone combined, naloxone and atropine 13 ?g/kg/hr, and morphine and bethanechol 16 ?g/kg/hr. Naloxone caused a significant increase in amylase outputs during the basal period and in response to all stimuli, without affecting bicarbonate outputs. This effect was blocked by atropine. Morphine caused significant reductions in both amylase and bicarbonate outputs with all stimuli. This was accompanied by a reduction in amylase concentrations but not in those of bicarbonate. Bethanechol counteracted these effects of morphine. The combination of morphine and naloxone had no significant effect on pancreatic secretion. Changes in pancreatic polypeptide levels did not explain the observations. In conclusion, these studies suggest that endogenous opioids exert a modulating influence on human pancreatic secretion. Morphine has a significant inhibitory effect on enzyme output as in animal studies. These effects may be mediated through an effect on acetylcholine release from nerve terminals and an interaction with cholecystokinin.