A brain to blood carrier-mediated transport system for arginine vasopressin (AVP) was investigated in mice after intraventricular injection of iodinated AVP and varying amounts of unlabeled material or candidate inhibitors. Residual activity in the brain detected after decapitation was used as the main determinant of transport activity. The half-time disappearance of iodinated AVP from the brain was 12.4 min, the Vmax was 1.41 nmol/g-min, and the apparent Km was 28.7 nmol/g. A 30-nmol dose of AVP, mesotocin, arginine vasotocin, pressinoic amide, pressinoic acid, tocinoic acid, and lysine vasotocin, but not oxytocin, lysine vasopressin, AVP free acid, tocinoic amide, Tyr-MIF-1, or cyclo Leu-Gly, significantly (P less than 0.05) inhibited the transport of iodinated AVP out of the brain. The 30 nmol dose of AVP had no effect on the transport of iodide or iodotyrosine out of the brain. High-performance liquid chromatography showed that 59.2% of the radioactivity found in the blood 2 min after an i.c.v. injection of labeled AVP eluted at the same position as labeled AVP compared with 68.8% of radioactivity eluting at that position after material was infused i.v. for 2 min. This indicates that intact peptide is transported across the blood-brain barrier and that most of the degradation of AVP occurs during circulation in the blood. Calculations based on the appearance of radioactivity in the periphery showed that 56.2% of the material injected centrally would have been transported into the periphery by 10 min. This appearance of material in the periphery was inhibited by the simultaneous injection of an excess of unlabeled peptide. Water loading significantly decreased the brain to blood transport rate of AVP by 40%. It is concluded that a saturable system exists for brain to blood transport of AVP and some structurally similar peptides.