Conscious rabbits underwent six 4-min occlusion and 4-min reperfusion cycles for 3 consecutive days (day 1, 2, and 3); on day 1, rabbits received intravenous vehicle [preconditioning (PC)] (group I, n = 6), superoxide dismutase (SOD; group II, n = 5), catalase (group III, n = 6), or the hydroxyl radical (. OH) and peroxynitrite (ONOO-)) scavenger N-2-mercaptopropionyl glycine (MPG [group IV], n = 6). In the PC group, the recovery of systolic wall thickening (WTh) after the sixth reperfusion was markedly improved on days 2 and 3 compared with day 1 and the total deficit of WTh was correspondingly reduced, indicating a late PC effect against myocardial stunning. Neither SOD nor catalase had any significant effect on the severity of stunning on day 1 or on the development of late PC on days 2 and 3, despite high plasma levels. In contrast, MPG markedly attenuated the severity of stunning on day 1 and prevented the development of late PC on day 2. Two additional groups of rabbits received an intracoronary infusion of vehicle (group V, n = 4) or the reactive oxygen species (ROS) generating solution [cumene hydroperoxide (CuOOH, group VI, n = 7)] on day 0, and were then subjected to the six occlusion/reperfusion cycles on days 1, 2, and 3. In group VI, infusion of CuOOH elicited a late PC effect 24 h later (on day 1). Taken together, these results demonstrate that oxidant species play an essential role in triggering the development of late PC against stunning in conscious rabbits. The fact that late PC was blocked by MPG and mimicked by CuOOH implicate ONOO- and/or .OH as the oxygen species responsible for the initiation of this phenomenon. In addition, the finding that exogenous ROS (CuOOH) reproduced the phenotype of late PC indicates that ROS are not only necessary but also sufficient to trigger this defensive adaptation of the heart to stress.