The effects of 10 differently structured bile acids on bile flow and composition were studied in anesthetized, bile duct-cannulated guinea pigs. At the infusion rates of 2 and 4 mumole/min/kg, all bile acids produced choleresis. The most potent was chenodeoxycholate, which increased bile flow by an average of 31.25 microliters/mumole of bile acids excreted in bile. The weakest choleretic was tauroursodeoxycholate (11.02 mu/mumole). When the choleretic activity was plotted against bile acid hydrophobicity (high-performance liquid chromatography retention factor, obtained from the literature), linearity was observed with similarly conjugated bile acids. The order of potency was deoxycholate greater than chenodeoxycholate greater than cholate greater than ursodeoxycholate, both for the glycine and taurine conjugates, and for the unconjugated bile acids as well. Conjugation was also important, and the rank ordering for the choleretic activity (unconjugated bile acids greater than glycine-conjugates greater than taurine-conjugates) was the same as that for the hydrophobicity. When the choleretic activity was plotted against bile acid micellar aggregation number (in 0.15 M NaCl at 36 degrees C, obtained from the literature), a linear, direct relationship was observed. All bile acids produced similar effects on bile electrolyte concentrations: both bicarbonate and chloride slightly declined during choleresis, whereas bile acid concentrations increased. These studies suggest that, in the guinea pig the differing choleretic activities of differently structured bile acids are not due to their forming micelles in bile of different sizes; either the more hydrophobic bile acids form vesicles, whereas the more hydrophilic form micelles; or bile acids produce choleresis, in part or exclusively, by stimulating an additional secretory mechanism, possibly an inorganic ion pump; or both.