1. Hepatic removal and metabolism as well as biliary excretion of noradrenaline were studied. Rat livers were perfused in situ for 60 min with Krebs-Henseleit buffer at 37 degrees C containing 2 nM [3H]-(-)-noradrenaline. [3H]-noradrenaline and its [3H]-metabolites were determined in liver, venous effluent and bile. 2. Removal of [3H]-noradrenaline by the liver, calculated as the sum of total radioactivity in the liver at the end of perfusion plus total radioactivity in the bile formed during perfusion plus [3H]-metabolites in the venous effluent formed during perfusion, was 40.2 +/- 6.9 pmol g-1 h-1. This removal corresponded to about 25% of the amount of [3H]-noradrenaline offered to the liver. 3. A proportion of the [3H]-noradrenaline (86.8%) taken up by the liver was metabolized, 13.2% remained unmetabolized in the liver and 0.019% was excreted unmetabolized into the bile. The most abundant metabolites were those present in the [3H]-OMDA fraction (72.5%), followed by [3H]-NMN (15.8%), [3H]-DOPEG (6.1%) and [3H]-DOMA (5.6%). Some of these metabolites (66.6%) were recovered from the venous effluent, 32.7% from the liver and only 1.3% from the bile. The amount of [3H]-noradrenaline present in the liver at the end of the perfusion produced a tissue:perfusion medium ratio of 2.6. 4. Simultaneous inhibition of monoamine oxidase and catechol-O-methyl transferase with pargyline (75 mg kg-1, i.p., 3 h before) and tolcapone (1 microM), respectively, markedly reduced the formation of [3H]-NMN, [3H]-DOPEG and [3H]-DOMA, but did not affect the hepatic removal of [3H]-noradrenaline, the content of [3H]-noradrenaline in the liver, the formation of [3H]-OMDA or the excretion of [3H]-noradrenaline and its [3H]-metabolites into the bile. 5. Treatment with an uptake2 blocker, corticosterone (40 microM), did not change the hepatic removal and metabolism of [3H]-noradrenaline or the biliary excretion of [3H]-noradrenaline and its [3H]-metabolites. 6. These findings indicate that the perfused rat liver efficiently removed and metabolized [3H]-noradrenaline, both monoamine oxidase and catechol-O-methyl transferase being involved in the metabolism of this amine. The apparent lack of effect of monoamine oxidase and catechol-O-methyl transferase inhibition on the formation of [3H]-OMDA may be due to the presence, especially in the liver, of conjugated metabolites of [3H]-noradrenaline in the [3H]-OMDA fraction. These results also show that uptake2 does not seem to be involved in the hepatic uptake of [3H]-noradrenaline, confirming previous findings. Finally, the results indicate that the rat liver perfused with Krebs-Henseleit buffer is not a suitable experimental model for studies on the biliary excretion of catecholamines.