Previous studies have demonstrated that adenosine significantly enhances myocardial salvage after 90 minutes of regional ischemia. To determine its effect after prolonged ischemia, closed-chest dogs underwent 3 hours of left anterior descending artery occlusion followed by 72 hours of reperfusion. Intracoronary adenosine (3.75 mg/min; at 1.5 ml/min:total volume = 90 ml; n = 10) or an equivalent volume of saline (1.5 ml/min: total volume = 90 ml; n = 9) was infused into the left main coronary artery during the first 60 minutes of reperfusion. Regional myocardial blood flow was assessed serially with microspheres and regional ventricular function was assessed by contrast ventriculography. Infarct size was determined histologically. Light and electron microscopy were utilized to assess neutrophil infiltration and microvascular injury. Adenosine failed to reduce infarct size expressed as a percentage of the area at risk (38.0 ± 4.9% versus 34.8 ± 4.6%; p = NS) or to improve regional ventricular function as measured by the radial shortening method (3.2 ± 1.8% versus 2.2 ± 3.1%; p = NS) at 72 hours after reperfusion. Vasodilatory effects were not observed in the endo- and midmyocardial regions of the ischemic zone during adenosine administration. This was associated with a similar extent of capillary endothelial changes and neutrophil infiltration in both adenosine-treated and saline control groups. These results suggest that severe functional abnormalities are present in the vasculature after 3 hours of ischemia and that adenosine therapy is ineffective in enhancing myocardial salvage.