BACKGROUND: It has been reported that one or more intermittent reperfusion(s) during ischemia may be beneficial to the myocardium by washing out catabolites that have accumlated during ischemia. We evaluated the effect of four cycles of ichemia (2 minutes) and reperfusion (3 miutes), i.e., preconditioning on sustained ischemia (20 minutes) and reperfusion (60 minutes) using isolated Langendorff-perfused rabbit hearts. METHODS: After a fifty-minutes recovery phase, LVP , dLVP/dt and ECG were simultaneously recorded and ultrastructure of the stunned(or risk) area of the left ventricle was examined with conventional methods. RESULTS: In the preconditioned hearts, functional parameters such as LVPP(peak pressure), LVPP recovery rate and LVEDP(end-diastolic pressure) reached to 99.6+/-4.38mmHg, 98.0+/-4.67% and 14.0+/-2.90mmHg (109.3+/-2.91mmHg, 109.4+/-1.29mmHg and 10.7+/-2.67mmHg for the controls), respectively, after 30 minutes from the onset of reperfusion and maintained as in the controls(p>0.01). In contrast, in the ischemia-reperfusion hearts, LVPP and LVPP recovery rate were significantly reduced(81.6+/-6.83mmHg and 85.7+/-5.30%;p<0.05) and LVEDP elevated(21.2+/-3.00mmHg) but dP/dtmax, heart rate and ECG patterns were not significantly different between the preconditioned and the ischemia-refusion hearts during reperfusion. Furthermore, irreversible myocardial injury was homogeneous(both subendo- and subepicardial) in the ischmia-reperfusion hearts but only focal(subendocardial) in preconditioned hearts. CONCLUSION: These results suggest that preconditioning induced by very short periods of ischemia and reperfusion may enhance recovery of the left ventricular function and delay ultrastructhral changes to a certain extent during reperfusion.