Combined administration of buthionine sulfoximine (2.5 mmol/kg, i.p.) and diethyl maleate (1.0 ml/kg, i.p.) resulted in a near-complete depletion of hepatic glutathione (0.02 mumol/g liver vs 5.17 mumol/g in saline-treated controls) in male Sprague-Dawley rats. Bile flow was markedly reduced in the rats as compared with the controls and glutathione was not detected in the bile. The linear regression of the correlation between bile flow and endogenous bile-acid excretion rates revealed that no bile acid-independent bile flow was produced in the glutathione-depleted rats. The bile flow was partially restored by an intravenous infusion of glutathione isopropyl ester (1.17 mmol/kg/hr). Glutathione levels were increased in the bile (16 nmol/kg/min) and in the liver (0.55 mumol/g) at the end of the 100 min infusion period of the ester. The increments in bile flow rates were not proportional to the biliary excretion rates of bile acids or glutathione, and the flow rates suddenly increased when glutathione levels in the bile reached an apparent threshold. The increments, not accompanied with an excretion of diethyl maleate-glutathione conjugate, were much greater than expected from the osmotic choleresis of glutathione in the bile. These results indicate that hepatic glutathione above a certain level is required for the formation of a portion of bile flow, and that an intravenous administration of glutathione isopropyl ester is effective in partially restoring the bile formation impaired by glutathione depletion.