Hearts of anesthetized rats (n = 12 in each of 7 groups) were subjected to 5 min of regional ischemia; on reperfusion, 100% exhibited ventricular tachycardia (VT), 83% ventricular fibrillation (VF), and the mean total number of premature ventricular complexes (PVC) was 681 +/- 220. Hearts were then reverted to sinus rhythm and allowed to "recover" for 10, 20, 30, 60, 120 min, 1 or 3 days; they were again subjected to 5 min of ischemia and reperfusion. With recovery periods of 10 and 20 min a second episode of ischemia and reperfusion resulted in very few arrhythmias: 0% VF in both instances, 17 and 8% VT, and only 4 +/- 3 and 9 +/- 5 PVC, respectively. As the recovery period increased to 3 days there was a progressive return of vulnerability to reperfusion-induced arrhythmias. However, this return was slow, such that with a recovery time of 60 min, only 33% of hearts fibrillated. In additional studies, hearts were subjected to 0.5, 1.0, 3.0, or 5 min of regional ischemia, 10 min of reperfusion, 5 min of ischemia, and a second period of reperfusion. The vulnerability to arrhythmias during the second period of reperfusion was found to correlate inversely with the incidence of arrhythmias elicited by the first episode of reperfusion. We propose that a very short period of ischemia and reperfusion "preconditions" the myocardium so as to influence the vulnerability of the heart to subsequent reperfusion-induced arrhythmias, and that "recovery" from this effect can take a substantial period of time.