There are only a few ligands available for labelling brain receptors simply because in vivo binding requires more severe experimental conditions than in vitro binding. We now describe the in vivo binding properties of [H-3] RP 62203, a new potent and selective 5-HT2 antagonist. After intravenous injection into rats, [H-3]RP 62203 accumulated predominantly in brain regions containing 5-HT2 receptors, with a frontal cortex/cerebellum ratio of 6 to 7. A good correlation was obtained between the regional distribution of [H-3]RP 62203 in the brain and the density of 5-HT2 receptors measured in vitro. In vivo binding of [H-3] RP 62203 was saturable in the frontal cortex but not in the cerebellum. The B(max) in the frontal cortex was equal to 42.5 fmol/mg, thus in the same range as was found in vitro. The 5-HT2 selectivity was ascertained by displacement (prevention) experiments; 5-HT2 antagonists or the agonist 2,5-dimethoxy-4-iodophenylisopropylamine could prevent specific labelling of [H-3]RP 62203 only in brain regions containing 5-HT2 receptors. Interestingly, the radioactivity remaining in various brain regions after displacement with pipamperone corresponded exactly to that measured in the cerebellum, with or without pipamperone. In conclusion, [H-3]RP 62203 possesses striking properties of in vivo binding which make it a suitable candidate for examining 5-HT2 receptors in human brain by positron emission tomography scanning.