In an attempt to characterize in vivo the central serotonergic (5-HT2) receptors in humans with positron emission tomography (PET), we have used 11C-labeled-ketanserin, a seratonergic antagonist that has high affinity and selectivity for the 5-HT2 receptors in vitro. Earlier in vivo studies in rats had demonstrated a preferential accumulation of 3H-ketanserin in the frontal cortex relative to cerebellum, in accordance with known differences in density of 5-HT2 receptor in these two brain structures (5-HT2 receptors are dense in frontal cortex and sparse or absent in cerebellum). In rats, tracer accumulation in frontal cortex represented specific binding of 3H-ketanserin to 5-HT2 receptors in vivo as demonstrated by inhibition, saturation, and displacement studies. In 5 control subjects, we found a statistically significant retention of 11C-ketanserin in frontal cortex relative to cerebellum after intravenous injection of a tracer dose of the radioligand, suggesting specific in vivo binding of 11C-ketanserin to frontal cortex. To substantiate this hypothesis, we studied 4 subjects administered unlabeled chlorpromazine (CPZ) intramuscularly in therapeutic amounts (75 mg) two hours before the PET study. Pretreatment with CPZ decreased significantly the retention of 11C-ketanserin by the frontal cortex, indicating that almost total occupation of the ketanserin receptors by CPZ had been achieved prior to the injection of the radioligand. Despite these positive results, both the duration and the magnitude of tracer retention by frontal cortex in control subjects were much smaller than was reported in rats, presumably indicating lower specific binding, higher nonspecific binding, and faster drug metabolism in humans.(ABSTRACT TRUNCATED AT 250 WORDS)