Biliary bicarbonate secretion may occur by transport of bicarbonate itself or of H+ (or OH-). To distinguish between these two mechanisms, we have studied the effects of bicarbonate deprivation or substitution by weak acids in the perfusate of isolated rat livers on ursodeoxycholate-induced bicarbonate secretion. Livers were perfused with an erythrocyte-free solution containing either the impermeant buffer Tricine (25 mM) or 25 mM Tricine and 13 mM bicarbonate, acetate, or 5,5-dimethyloxazolidine-2,4-dione (DMO), and ursodeoxycholate was infused. Tauroursodeoxycholate, which does not stimulate bicarbonate secretion, served as a control. During ursodeoxycholate infusion 1) the increase in bile flow, in microliter X min-1 X g liver-1 (+/- SE), was significantly higher in livers perfused with Tricine and bicarbonate (1.29 +/- 0.06), Tricine and acetate (1.46 +/- 0.07), and Tricine and DMO (1.30 +/- 0.04) than in livers perfused with Tricine alone (0.99 +/- 0.04); and 2) biliary bicarbonate, acetate, or DMO concentrations and bile pH were significantly higher than the corresponding perfusate values. In contrast, during tauroursodeoxycholate infusion bile flow was the same whatever the perfusate, and bile pH was lower than pH of the perfusate. Therefore, ursodeoxycholate-induced choleresis and bile alkalinization do not depend on bicarbonate as such (which can be replaced by acetate or DMO). This suggests that ursodeoxycholate-induced biliary bicarbonate secretion is the result of H+ (or OH-) transport rather than transport of bicarbonate itself.