1 The e ect of the administration of pertussis toxin (PTX) as well as modulators of di erent subtypes of K + channels on the antinociception induced by clonidine and guanabenz was evaluated in the mouse hot plate test. 2 Pretreatment with pertussis toxin (0.25 mg per mouse i.c.v.) 7 days before the hot-plate test, prevented the antinociception induced by both clonidine (0.08 ± 0.2 mg kg 71 , s.c.) and guanabenz (0.1 ± 0.5 mg kg 71 , s.c.). 3 The administration of the K ATP channel openers minoxidil (10 mg per mouse, i.c.v.), pinacidil (25 mg per mouse, i.c.v.) and diazoxide (100 mg kg 71 , p.o.) potentiated the antinociception produced by clonidine and guanabenz whereas the K ATP channel blocker gliquidone (6 mg per mouse, i.c.v.) prevented the a 2 adrenoceptor agonist-induced analgesia. 4 Pretreatment with an antisense oligonucleotide (aODN) to mKv1.1, a voltage-gated K + channel, at the dose of 2.0 nmol per single i.c.v. injection, prevented the antinociception induced by both clonidine and guanabenz in comparison with degenerate oligonucleotide (dODN)-treated mice. 5 The administration of the Ca 2+ -gated K + channel blocker apamin (0.5 ± 2.0 ng per mouse, i.c.v.) never modi®ed clonidine and guanabenz analgesia. 6 At the highest e ective doses, none of the drugs used modi®ed animals' gross behaviour nor impaired motor coordination, as revealed by the rota-rod test.7 The present data demonstrate that both K ATP and mKv1.1 K + channels represent an important step in the transduction mechanism underlying central antinociception induced by activation of a 2 adrenoceptors.