Preconditioning myocardium with brief episodes of ischemia reduces energy demand and delays cell death during a subsequent ischemic episode. We hypothesized that postischemic contractile dysfunction after the brief ischemic episodes ("stunning") causes this reduced energy demand. If this hypothesis is correct, then cardioprotection should persist as long as mechanical function still is depressed at the onset of sustained ischemia. To analyze the temporal relationship between preconditioning and stunning, infarct size was compared in two groups of open-chest anesthetized dogs that were preconditioned with a 15-min coronary occlusion followed by a sustained 40-min occlusion. One group received 5 min of reperfusion and the second group received 120 min of reperfusion between occlusions. Nonpreconditioned controls received a single 40-min occlusion. A 15-min occlusion caused severe stunning, which did not improve during 2 h of reperfusion. In the 5-min reflow group, preconditioning resulted in dramatically smaller infarcts, averaging 2.2 +/- 0.9% of the area at risk vs. 26.5 +/- 4.2% in controls (P less than 0.01), confirmed by a marked shift in the inverse relationship between collateral blood flow and infarct size. Despite persistently severe stunning in the 120-min reflow group, infarct size was intermediate, averaging 12.3 +/- 2.7% (P less than 0.05 vs. 5-min reflow; P less than 0.01 vs. control), and the infarct vs. flow regression had returned toward control. Thus the cardioprotective effect of preconditioning was attenuated when the intervening reperfusion time was extended, even though severe contractile dysfunction persisted. We conclude that myocardial stunning, per se, is insufficient to cause preconditioning.