The effects of intrahepatic portal infusion of hyperosmotic substances or angiotensin II (A II) on the plasma arginine vasopressin (pAVP) concentration were investigated. At the beginning of each experiment, rats anesthetized with sodium pentobarbital were infused with 0.9% NaCl at a rate of 0.037 ml/min via the hepatic portal vein for 30 min. Subsequent intraportal infusions of 1.4, 1.8 and 2.7% NaCl for 5 min at a rate of 0.052 ml/min raised pAVP levels from 15.6 +/- 5.2 pg/ml to 26.0 +/- 1.7, 61.5 +/- 8.8 and 37.5 +/- 6.6 pg/ml, respectively. However, subsequent intraportal infusions of 0.9 or 3.6% NaCl and intrajugular infusions of 1.8 or 3.6% NaCl did not further raise the pAVP levels. Intraportal infusions of 21 and 42% sucrose raised pAVP levels to 28.3 +/- 2.9 and 24 +/- 2.0 pg/ml, respectively, whereas glucose did not change the pAVP concentration. When A II was infused via the intraportal vessel at a rate of 30 ng/kg/min for 10, 15 or 30 min, pAVP levels became 29.3 +/- 2.7, 32.1 +/- 4.0 or 25.8 +/- 3.2 pg/kg/min respectively. PAVP levels did not change when the same dose of A II was infused into the jugular vein. The increase of pAVP induced by A II was prevented when angiotensin II analogue (A II A), 1Sar, 8Ileu-A II, was infused into portal vein simultaneously. These results not only suggest that the osmoreceptors exist in the area located somewhere within the hepatic portal vascular bed, but also that the osmoreceptors are somewhat different in nature from those of hypothalamic osmoreceptors. Furthermore, these results suggest that A II in the hepatic portal vein plays a part in controlling AVP release.