The temporal properties of kappa-opioid receptor (kappa-OR) mediated cardioprotection are less well characterised than delta-opioid receptor (delta-OR) responses. This study was aimed at delineating the time course of kappa-OR-mediated protection in two experimental models: an in vivo rat model of regional myocardial infarction (30 min of left coronary artery occlusion with 120 min of reperfusion), and an in vitro perfused murine heart model (undergoing 25 min of global ischemia and 45 min of reperfusion). In the rat model, the selective kappa-OR agonist U50, 488 (0.1 mg/kg, IV bolus), administered either 10 min prior to ischemia or 5 min prior to reperfusion, significantly reduced infarct size (38 +/- 3% and 43 +/- 2% infarct size/area-at-risk (IS/AAR), respectively; P < 0.05) compared to untreated rats (56 +/- 1% IS/AAR). Administration of U50, 488 10 s after onset of reperfusion failed to elicit protection. Cardioprotection with U50,448 administered immediately prior to reperfusion was abolished by a kappa-OR antagonist, (0.1 mg/kg nor-BNI), given 10 min prior to reperfusion. In the in vitro murine model, untreated hearts exhibited 28 +/- 2% (IS/AAR) infarct size. Infusion of U50, 488 (at a final 100 nM concentration) significantly limited infarct size in mouse hearts when applied at the onset of reperfusion (15 +/- 2% IS/AAR; P < 0.05), yet failed to afford protection when infused prior to ischemia. Additionally, in both models studied, treatment with either wortmannin or 5-hydroxydecanoate (5-HD) abrogated the protective effects of U50,488 applied just prior to reperfusion. In summary, kappa-ORs afford cardioprotection primarily when activated prior to and not after reperfusion. This protection may involve activation of the PI3 kinase (PI3K) pathway and mitochondrial (mito) K (ATP) channels.