The aim of this study was to measure the role of the duodenal loop in biliopancreatic secretion in man by infusing various stimuli at the ampulla of Vater and collecting duodenal contents at the ligament of Treitz, above an occluding balloon. Perfusion at 10 ml/min of a first mixture of aminoacids - phenylalanine (47.2 mmol), methionine (38.2 mmol), tryptophan (11 mmol), valine (61.6 mmol) - increased cholecystokinin (CCK) plasma concentrations and duodenal bile salt output (p less than 0.005) as compared with a control electrolyte solution, but did not change pancreatic enzyme secretion significantly; duodenal infusion of another aminoacid mixture - arginine (32.4 mmol), histidine (14.1 mmol), leucine (36 mmol), isoleucine (21.5 mmol), lysine (31 mmol), threonine (23 mmol) - did not change CCK plasma concentrations, bile salt or pancreatic enzyme output. The respective role of duodenal distension and endogenous CCK was investigated by perfusing the first aminoacid solution and the control solution at 2, 5, and 10 ml/min. Changing the perfusion rate of control solution from 2 to 5 ml/min led to a significant increase (p less than 0.01) in pancreatic secretion with no further increase at 10 ml/min. Bile salt output was not influenced by the perfusion rate of control solution. During the perfusion of the aminoacid solution, despite a stepwise increase in CCK release, the only significant change in pancreatic secretion was an increase of lipase output (p less than 0.05) when the infusion rate was raised from 2 to 5 ml/min. Our results suggest that duodenal CCK release (1) depends on the nature of aminoacids (2) has predominant role in the regulation of pancreatic secretion at low perfusion rate but is less effective when superimposed on a mechanical stimulus caused by duodenal distension (3) is a major stimulus for gall bladder contraction which is not influenced by duodenal distension.