To characterize the transport mechanisms responsible for formation of canalicular bile, we have examined the effects of ion substitution on bile acid-dependent and bile acid-independent bile formation by the isolated perfused rat liver. Complete replacement of perfusate sodium with choline and lithium abolished taurocholate-induced choleresis and reduced biliary taurocholate output by greater than 70%. Partial replacement of perfusate sodium (25 of 128 mM) by choline reduced bile acid-independent bile formation by 30% and replacement of the remaining sodium (103 mM) by choline reduced bile acid-independent bile formation by an additional 64%. In contrast, replacement of the remaining sodium (103 mM) by lithium reduced bile acid-independent bile formation by only an additional 20%, while complete replacement of sodium (128 mM) by lithium reduced bile formation by only 17%, and lithium replaced sodium as the predominant biliary cation. Replacement of perfusate bicarbonate by Tricine, a zwitterionic amino acid buffer, decreased bile acid-independent bile formation by greater than or equal to 50% and decreased biliary bicarbonate output by approximately 60%, regardless of the accompanying cation. In separate experiments, replacement of sodium by lithium essentially abolished Na,K-ATPase activity measured either as ouabain-suppressible ATP hydrolysis in rat liver or kidney homogenates, or as ouabain-suppressible 86Rb uptake by cultured rat hepatocytes. These studies indicate that bile acid(taurocholate)-dependent bile formation by rat liver exhibits a specific requirement for sodium, a finding probably attributable to the role(s) of sodium in hepatic sodium-coupled taurocholate uptake and/or in maintenance of Na,K-ATPase activity. The surprising finding that bile acid-independent bile formation was substantially unaltered by complete replacement of sodium with the permeant cation lithium does not appear to be explained by Na,K-ATPase-mediated lithium transport. Although alternative interpretations exist, this observation is consistent with the hypothesis that much of basal bile acid-independent bile formation is attributable to an ion pump other than Na,K-ATPase, which directly or indirectly mediates bicarbonate transport.