Objective The hypothesis that naloxone blockade of ischemic preconditioning (IP)-induced infarct limitation does not require central nervous system participation was evaluated using quaternary naloxone in anesthetized rabbits (Study I) and naloxone hydrochloride in isolated rabbit hearts (Study II). Methods In Study I, rabbits underwent 30 min coronary artery occlusion and 180 min reperfusion. IP was elicited with a 5 min coronary artery occlusion beginning 15 min before the 30 min occlusion. Intravenous naloxone methiodide, 12.9 mg/kg, was bolused 10 or 1 min before IF. In Study II, rabbit hearts underwent 15 min coronary artery occlusion and 120 min reperfusion. IP was elicited with 2 cycles of 5 min coronary artery occlusion plus 5 min reperfusion, beginning 20 min before the 45 min occlusion. Naloxone hydrochloride, 1 mu mol/L or 100 mu mol/L, was added to the buffer perfusate for 25 min preceding the long coronary artery occlusion. In both studies, infarct size was assessed with tetrazolium, normalized to risk volume, and analyzed using ANOVA. Results In both studies, IP reduced infarct size compared to control (6.3 +/- 2.3 vs. 29.5 +/- 4.4, P = 0.007, Study I; 11.8 +/- 4.7 vs. 47.7 +/- 6.7, P = 0.03, Study II). In Study I, IP was not blocked when naloxone methiodide was given 10 min before IP (13.8 +/- 1.8 vs. 42.3 +/- 5.4, P = 0.004) but was blocked when given 1 min before IP (25.3 +/- 7.2 vs. 28.4 +/- 5.0, Pr ns). In Study II, infarct size was intermediate in the 1 mu mol/L naloxone hydrochloride+IP group (19.0 +/- 6.5 vs. 48.9 +/- 8.4, P = ns) but IP was blocked by 100 mu mol/L naloxone hydrochloride (62.6 +/- 4.5 vs. 56.2 +/- 6.7, P = ns). Conclusion Naloxone blockade of IP-induced infarct limitation involves a cardiac mechanism.