To determine if the mechanism of action of clinical and investigational antiepileptic and antimyoclonic drugs or neuropeptides involves direct actions at serotonin (5-HT) receptors, the activity of various compounds in vitro at 5-HT1 (with subtypes) and 5-HT2 sites was measured in adult rat brainstem, spinal cord, and neocortex. Adrenocorticotropic hormone (ACTH1-39) noncompetitively inhibited specific binding at 5-HT1, 5-HT1A, and 5-HT2 sites in brainstem and neocortex [concentrations required to displace 50% of ligand binding (IC50S) 4-8 X 10(-5) M]. ACTH1-24, ACTH1-17, and ACTH4-10 were sequentially less active, and ACTH34-39 and corticosterone were inactive. D-Ala2, Leu5-enkephalinamide, but not D-Ala2, Met5-enkephalinamide, also displaced spinal and neocortical 5-HT2 sites (IC50 6 X 10(-5) M). Piracetam, glycine, and the clinical antiepileptics valproate, phenacemide, phenytoin, carbamazepine, phenobarbital, diazepam, clonazepam, nitrazepam, and ethosuximide did not displace serotonergic radioligands, but melacimide showed some activity at 5-HT1 sites (IC50 7-9 X 10(-5) M). Anticonvulsant inactivity at 5-HT receptors in vitro correlates with the lack of antimyoclonic activity in 5-HT lesion myoclonic models but not with antimyoclonic efficacy in humans. These data indicate that acute effects of these anticonvulsants cannot be attributed to direct action at the 5-HT receptor recognition site in the rat. In contrast, ACTH showed mild in vitro displacement and regional specificity but only at micromolar concentrations.