Adenosine A, receptor stimulation increases myocardial resilence to ischaemia many hours later but the mechanism of protection is unclear. We hypothesized that A, receptor-induced delayed cardioprotection is mediated by Ii, channel opening. Rabbits were pretreated with saline or the selective A, receptor agonist 3-chloro-N-6-cyclopentyladenosine (CCPA), 0.1 mg/kg i.v. After 24h, they were anaesthetized and subjected to 30 min left coronary artery occlusion (CAO) and 120 min reperfusion. Twenty minutes before CAO, either glibenclamide 0.3 mg/kg, sodium 5-hydroxydecanoate (5-HD) 5 mg/kg, or a corresponding volume of vehicle solution were given i.v. Infarct size as a percentage of ischaemic risk volume (I/R%) was assessed by triphenyltetrazolium. In the absence of either glibenclamide or 5-HD, CCPA pretreatment resulted in significant limitation of infarction (I/R 23.4 +/- 3.3% vs 39.6+/-2.6% in saline pretreated animals, P<0.01). Administration of glibenclamide before CAO abolished this delayed protective effect of CCPA (I/R 37.4 +/- 4.7%), as did 5-HD (I/R 48.8 +/- 3.7%). Neither glibenclamide nor 5-HD significantly modified I/R in saline pretreated animals. Risk volume was similar in all groups (0.8-1.1cm(3)). Systemic haemodynamic variables were not markedly different between any of the groups immediately before or during the ischaemia-reperfusion protocol. Thus, delayed protection following transient A, receptor activation was abolished by pre-ischaemic treatment with either glibenclamide or 5-HD, providing pharmacological evidence that Ii, channel opening mediates A, agonist induced delayed myocardial protection. The inhibitory effect of 5-HD suggests specific involvement of mitochondrial K-ATP but the mechanisms of sub-acute regulation of K-ATP function following A(1) receptor stimulation remain to be determined. (C) 1999 Academic Press.