Because myocardial reperfusion injury may be caused by various blood constituents, a transient period of blood-free reperfusion was evaluated in closed chest dogs subjected to a 90 min angioplasty balloon occlusion of the left anterior descending coronary artery. In the treated group (n = 13), the balloon remained inflated for an additional 15 min while the infarct vessel was perfused with an acellular oxygenated perfluorochemical emulsion (Fluosol). The balloon was then deflated, permitting blood reperfusion. In the control group (n = 13), the balloon was deflated after 90 min of coronary occlusion. One week after infarction, the area at risk was defined in vivo by monastral blue dye staining, and the area of myocardial necrosis was assessed using triphenyltetrazolium chloride staining with histologic confirmation. Major determinants of infarct size, including rate-pressure product, area at risk and severity of myocardial ischemia (assessed by the extent of ST segment elevation during coronary occlusion), were not significantly different in the two groups. Treated dogs demonstrated a 47% reduction in infarct size expressed as a percent of the area at risk compared with control dogs (27.0 +/- 4.4% versus 50.8 +/- 4.4%, p less than 0.01). Treated dogs also demonstrated a superior global left ventricular ejection fraction (57.5 +/- 2.5% versus 51.0 +/- 2.2%, p less than 0.05) and anterolateral (regional) ejection fraction (32.6 +/- 3.6% versus 19.8 +/- 3.9%, p less than 0.05) compared with values in control dogs assessed by contrast ventriculography after 1 week of reperfusion. It is concluded that a transient period of blood-free reperfusion with an oxygenated perfluorochemical reduces reperfusion injury in a canine model of myocardial infarction.