1 The mechanism underlying the anticataleptic properties of the atypical neuroleptic agent, clozapine, has been investigated in the rat. 2 The close structural analogues of clozapine, loxapine (0.1 mg kg 71 s.c.) and iso-clozapine (1 and 3 mg kg 71 s.c.) induced catalepsy in rats. In contrast, clozapine and the regio-isomer of loxapine, isoloxapine (up to 10 mg kg 71 s.c.) did not produce catalepsy, but at a dose of 1 mg kg 71 signi®cantly inhibited catalepsy induced by loxapine (0.3 mg kg 71 s.c.). 3 Radioligand binding assays showed that cataleptogenic potential was most clearly predicted by the D 2 /5-HT 1A , D 2 /5-HT 1B/1D and D 2 /a 2 -receptor a nity (K D ) ratios: i.e. 30 ± 100-fold higher ratios were calculated for loxapine and iso-clozapine, whereas the ratios were less than 1 for clozapine and isoloxapine. The ratios of a nities for D 2 to 5-HT 2A , 5-HT 2C or D 1 did not re¯ect the grouping of cataleptic and non-cataleptic compounds. 4 Co-treatment with the a 2 -adrenoceptor antagonists, yohimbine (1 ± 10 mg kg 71 s.c.), RX 821002 (1 ± 10 mg kg 71 s.c.) and MK-912 (0.3 and 1 mg kg 71 s.c.) dose-dependently inhibited the cataleptic response to loxapine (0.3 mg kg 71 ). Yohimbine (1 ± 10 mg kg 71 s.c.) also dose-dependently inhibited the cateleptic response to haloperidol (0.3 mg kg 71 s.c.). The a 2 -adrenoceptor antagonists had no e ect per se. 5 Neither yohimbine (10 mg kg 71 ) nor RX821002 (3 mg kg 71 ) altered the cataleptic response to the D 1 receptor antagonist, SCH 23390 (1 mg kg 71 s.c.), while, like clozapine, both compounds abolished the response to the 5-HT 2A receptor antagonist, MDL 100,151 (3 mg kg 71 s.c.). 6 The present data strongly implicate a 2 -adrenoceptor blockade in the anticataleptic properties of clozapine and suggest that its lack of extrapyramidal side e ects in the clinic may also be a consequence of this property.