Objective: ATP sensitive potassium (KATP) channels have been implicated in the mechanism of ischaemic preconditioning, though apparently not in the pentobarbitone anaesthetised rabbit model. The aim of this study was to test whether potassium channel activation and blockade would alter protection in ketamine-xylazine anaesthetised rabbits. Methods: In situ rabbit hearts (n = 50) received 30 min regional ischaemia and 3 h reperfusion. Some hearts were preconditioned by 5 min regional ischaemia and 10 min reperfusion prior to the long ischaemia. Infarct size was determined by tetrazolium staining. Results: In rabbits anaesthetised with ketamine-xylazine, brief preconditioning ischaemia continued to produce much smaller infarcts than in non-preconditioned animals [19(SEM 2)% v47(3)%, p ≤ 0.05]. Blocking KATP channels by pretreating with glibenclamide resulted in 35(3)% infarction in non-preconditioned hearts and aborted protection in preconditioned hearts [35(4)% infarction]. Substituting the potassium channel activator pinacidil for the short ischaemia caused comparable reductions in infarct size [28(4)%, p ≤ 0.05 v non-preconditioned hearts]. This protection, however, could be blocked by concomitant administration of the adenosine receptor blocker 8-(p-sulphophenyl)theophylline (SPT) [44(3)% infarction]. Conclusions: When ketamine-xylazine anaesthesia was employed, the protective effects of ischaemic preconditioning in the rabbit heart could be blocked by glibenclamide, and pinacidil could mimic the protection of ischaemic preconditioning. Because the protection afforded by pinacidil could be blocked by SPT, however, there is still some question whether the KATP channel is the end effector of preconditioning. Cardiovascular Research 1994;28:1337-1341