Basal and TRH-stimulated PRL levels have been reported to be increased in patients with severe liver disease. To elucidate the relative role of the liver and kidney in the elimination of PRL in such patients, plasma PRL levels were measured before and after TRH in peripheral, hepatic, and renal vein samples taken from eight patients undergoing diagnostic hepatic vein catheterization. In two patients, arterial PRL levels, effective renal plasma flow, and estimated hepatic plasma flow were measured. All patients had evidence of portal hypertension with parenchymal liver cell insufficiency varying from mild to severe. There was no significant difference among basal arterial, peripheral, hepatic, and renal venous PRL levels. After TRH administration, the mean renal PRL levels were significantly lower than the mean peripheral levels in six of eight individual patients and lower when mean renal venous PRL levels and mean peripheral levels of all patients taken together were compared (P < 0.001). The two patients in whom no significant difference was found were the only patients who also had evidence of disturbed kidney function. Mean hepatic venous levels after TRH stimulation were slightly lower than mean peripheral levels, although the difference was significant in only one patient. The difference between mean hepatic and peripheral venous PRL levels of all patients taken together was also significant (P < 0.025). In two patients, effective renal plasma flow measured 468 and 324 ml/ min, respectively, and estimated hepatic plasma flow measured 657 and 815 ml/min. From these data renal extraction of PRL was calculated to be 186.0 and 94.0 μg during the 60 min, while liver extraction of PRL amounted to only 16.4 and 11.4 μg, respectively. The finding that hepatic elimination was not markedly affected by the severity of liver disease suggests that the kidney is more important than the liver in the removal of PRL. In patients with liver disease the kidney is the main site of PRL elimination.