The authors tested the hypothesis that volatile anesthetic-induced preconditioning (APC) follows a similar time pattern as that described for ischemic preconditioning and that delayed APC is mediated by nitric oxide. A prospective randomized vehicle-controlled study. A university research laboratory. New Zealand white rabbits (n = 75). Rabbits were instrumented for the measurement of systemic hemodynamics and subjected to a 30-minute coronary artery occlusion (CAO) and 3 hours of reperfusion. Desflurane (1.0 minimum alveolar concentration) was administered for 30 minutes and was discontinued 0.5 hours, 2 hours, 3 hours, 12 hours, 24 hours, 48 hours, 72 hours, and 96 hours before CAO, respectively. In 2 separate experimental groups, the nitric oxide synthase inhibitor N-omega-nitro-L-arginine (L-NA) was administered 72 hours after the administration of 0.0 or 1.0 minimum alveolar concentration of desflurane. The infarct size was determined gravimetrically. Data are mean +/- standard deviation. Desflurane significantly (p < 0.05) reduced the infarct size compared with the control (63% +/- 12%, n = 7) when administered 0.5 hours (35% +/- 5%, n = 7), 2 hours (35% +/- 9%, n = 7), 24 hours (31% +/- 8%, n = 7), 48 hours (30% +/- 11%, n = 6), and 72 hours (39% +/- 5%, n = 6) before CAO. However, when desflurane was administered 3 hours (53% +/- 9%, n = 7), 12 hours (71% +/- 6%, n = 7), or 96 hours (66% +/- 5%, n = 7) before CAO, the myocardial infarct size was not reduced. The second window (72 hours) of preconditioning was abolished by the NOS inhibitor L-NA (52% +/- 16%, n = 7). L-NA alone had no effect on infarct size (64% +/- 11%, n = 7). Desflurane induces a first (0.5-2 hours) and second window of preconditioning (24-72 hours) in the rabbit model of acute myocardial infarction. The second window of APC is mediated by nitric oxide.