We sought to determine whether short periods of preconditioning ischemia limited infarct size in a model of low coronary collateral flow and whether the effects of preconditioning were transient or long-lasting. Rats underwent 90 minutes of coronary occlusion followed by reperfusion. In one group, the rats were preconditioned by three 3-minute occlusions, each separated by 5 minutes of reperfusion. The size of the myocardial infarcts were smaller in the preconditioned group (24.9 +/- 7.8% of the risk zone developed necrosis) versus the nonpreconditioned group (60.7 +/- 5.3%, p less than 0.01). In addition, the incidence of ventricular arrhythmias was reduced by preconditioning. However, when there was a delay of 1 hour or more between the brief episodes of preconditioning and the longer 90-minute occlusion, the beneficial effects of preconditioning were lost. Thus preconditioning reduced the volume of myocardium infarcted and the incidence of ventricular arrhythmias during a subsequent period of prolonged ischemia; however, these beneficial effects were transient.