Hepatic bile flow was measured and hepatic [(14C]mannitol clearance was calculated during depletion of the bile acid pool and during duodenal administration of cholic or chenodeoxycholic acid in 10 patients 7 to 12 days after operation for uncomplicated gallstone disease. The relationship between [(14)C]mannitol clearance and bile flow was linear, with a regression coefficient which was not significantly different from unity. This finding supported the assumption that mannitol clearance can be used as a measure of canalicular bile flow in man. Linear relationships between bile acid secretion rate and bile flow and bile acid secretion rate and [(14)C]mannitol clearance were found during depletion of the bile acid pool (secretion of cholic, chenodeoxycholic, and deoxycholic acids; relative concentrations, 1.0:1.0:0.5) as well as during cholic acid infusion (73 +/- 3% cholic acid in the secreted hepatic bile) and during chenodeoxycholic acid infusion (88 +/- 2% chenodeoxycholic acid in the secreted hepatic bile). The bile flow dependence on bile acids 0.011 +/- 0.002 ml mumoles(-1)) was not significantly different for "mixed" bile acid secretion, mainly cholic acid secretion, or mainly chenodeoxycholic acid secretion. Neither was the ductular bile flow (0.08 +/- 0.03 ml min(-1)) significantly influenced by the various experimental conditions. The canalicular flow--the bile acid-independent (0.17 +/- 0.05 ml min(-1)) and the bile acid-dependent flow--constituted 70 to 85% of the total bile flow. It is concluded that secretion of cholic acid and of chenodeoxycholic acid promote the same bile flow volume per micromole in man.