The respective role of central vs. peripheral CCK-B receptors in the recently reported anxiolytic effects of CCK-B antagonists remains to be firmly established. We therefore investigated the in vivo binding properties of cerebral CCK receptors after i.c.v. injection into mice of [3H]pBC 264 ([3H]propionyl-Tyr(SO3H)-gNle-mGly- Trp-(NMe)Nle-Asp-Phe-NH2), a highly potent, peptidase-resistant and selective CCK-B agonist. The specific binding of [3H]pBC 264 was reversible and saturable. The dose producing 50% receptor occupancy was 25 pmol and the Bmax was 0.9 pmol/brain 15 min after injection. I.c.v. administered CCK8 (ID50 8500 pmol) was 200-fold less potent than pBC 264 (ID50 43 pmol) in inhibiting specific [3H]pBC 264 binding; CCK8NS, CCK5 and CCK4 being slightly less potent than CCK8. Aminopeptidases play a major role in degrading CCK8 since the protected analog pCCK8 or CCK8 in the presence of an aminopeptidase inhibitor exhibited higher affinities than CCK8. I.v. administration of pBC 264 (20 mg/kg) inhibited [3H]pBC 264 specific binding by about 72%, confirming its ability to enter the brain. In contrast, CCK4 was unable to modify [3H]pBC 264 binding. As expected, the CCK-A antagonist (L364,718) did not inhibit [3H]pBC 264 binding, while at the highest dose used (40 mg/kg i.p.) the CCK-B antagonist (L365,260) inhibited binding by 20%. Several hypotheses are discussed to account for the very low i.v. doses of CCK4 and L365,260 needed to produce anxiogenic and anxiolytic responses, respectively.