The interoceptive stimulus induced by clozapine (5 mg/kg, i.p.) has been characterized in an operant drug discrimination procedure in the rat using a wide range of receptor subtype-selective agonists and antagonists. Only the muscarinic receptor antagonist scopolamine generalized fully to clozapine (> 80%). Partial generalization (defined here as 40% maximal generalization) was seen with the Di receptor antagonist SCH 23390 (43% maximal generalization), the α1-adrenoceptor antagonist prazosin (67%) and the α2-adrenoceptor antagonist methoxyidazoxan (42%). All other specific agents tested induced<25% maximal generalization, including the a2-adrenoceptor antagonist yohimbine (24%), the histamine H1receptor antagonist mepyramine (21%), the D2antagonist typical neuroleptic haloperidol (23%), the D4receptor antagonist L-745,870 (14%), the 5-hydroxytryptamine-lA (5-HT1A) receptor agonist S-14506 (8%), the 5-HT2Areceptor antagonists ketanserin (0%) and M100907 (12%), the 5-HT2B/2Creceptor antagonists SB 200646A (8%) and SDZ SER 082 (6%), and the 5-HT3receptor antagonist ondansetron (0%). The clozapine discriminative stimulus was not blocked by the dopamine D1receptor antagonist SCH 23390, or by the 5-HT1Areceptor antagonist WAY 100635, when given concomitantly with clozapine. Although the results suggest that muscarinic antagonism plays a major role in the clozapine cue, the results have to be considered in the light of the full generalization to clozapine seen with various antipsychotic agents which have very low affinity for muscarinic receptors, including zotepine, quetiapine, JL13 and PNU 96415 (a finding replicated in rats from the same breeding colony as those which generalized to scopolamine). Thus, generalization to clozapine for antipsychotics with multiple affinities but with low muscarinic affinity is probably mediated by additive or perhaps supra-additive actions at other receptors, although extensive studies with various combinations of drug mixtures are required to validate this hypothesis.