Myocardial preconditioning with brief coronary artery occlusions before a sustained ischemic period is reported to reduce infarct size. To determine the number of occlusive episodes required to produce the preconditioning effect, we performed single or multiple occlusions of the left circumflex coronary artery (LCx) followed by a sustained occlusion (60 minutes) of the LCx. Anesthetized dogs underwent one (P1), six (P6), or 12 (P12) 5-minute occlusions of the LCx. Each occlusion period was followed by a 10-minute reperfusion period. A 60-minute occlusion of the LCx followed the preconditioning sequences. A control group received a 60-minute occlusion of the LCx without preconditioning. All groups were subjected to 6 hours of reperfusion after which the heart was removed for calculating infarct size (IS), area at risk (AR), and left ventricular mass (LV). The IS/AR ratio for the control group was 29.8 +/- 4.4% (n = 17), which was substantially greater (p less than 0.001) than that of the preconditioned groups: P1, 3.9 +/- 1.3% (n = 14); P6, 0.4 +/- 0.3% (n = 5); and P12, 2.9 +/- 2.8% (n = 5). There were no significant differences in the IS/AR ratio among the three preconditioned groups. The AR/LV ratio was comparable among all groups and did not differ statistically: control, 40.4 +/- 1.3%; P1, 36.2 +/- 1.7%; P6, 36.1 +/- 1.7%; and P12, 37.3 +/- 2.1%. Collateral blood flow to the inner two thirds of the risk region determined with radiolabeled microspheres during ischemia did not differ significantly between the control group (0.03 +/- 0.01 ml/min/g, n = 8) and single occlusion group (0.06 +/- 0.02 ml/min/g, n = 8), indicating that the marked disparity in infarct size could not be attributed to differences in collateral blood flow. The data indicate that preconditioning with one brief ischemic interval is as effective as preconditioning with multiple ischemic periods.