Splanchnic uptake of haloperidol (HAL) and release of reduced haloperidol (RHAL) were studied in vivo in guinea pigs. Anesthetized animals with implanted cannulae in the aorta, the hepatic vein and the inferior vena cava were infused intravenously with HAL at a rate of 9.6 micrograms/min/animal for 90 min. Plasma HAL and RHAL in samples taken from the arterial and hepatic venous cannulae were measured by HPLC with an electrochemical detector. Contamination of the hepatic venous samples by blood from the inferior vena cava was ruled out by the validation method of tritiated water washout [Huang MT, J Appl Physiol 71: 359-364, 1991]. HAL concentrations plateaued at 70-80 ng/mL in the aorta and 5-7 ng/mL in the hepatic vein during the final 30 min of infusion. Splanchnic extraction of HAL was 91%. Hepatic blood flow was estimated to be 1.95 +/- 0.40 (SD) mL/min/g. If assuming that splanchnic uptake of HAL took place in the liver, a rate of uptake of HAL in the liver of 79.2 +/- 18.6 (SD) ng/min/g could be calculated by the Fick principle. The uptake in the whole liver accounted for 14% of the rate of HAL infusion into the animal. Plasma RHAL in the aorta, 6.4 +/- 6.6 (SD) ng/mL at 60 min and 9.4 +/- 4.6 (SD) ng/mL at the end of HAL infusion, was about 10-12-fold less than aortic HAL. The concentrations of RHAL in the hepatic vein were not significantly different from those in the aorta, indicating that splanchnic tissues including the liver are not responsible for plasma RHAL secretion. The highly efficient uptake of HAL as well as the ketone reductases found previously in vitro in liver microsomes of guinea pigs were probably involved only in biliary excretion of HAL.