Metformin improves cardiovascular outcomes in patients with type 2 diabetes compared with other glucose-lowering drugs. Experimental studies have shown that metformin can increase the intracellular concentration of adenosine monophosphate, which is a major determinant of the intracellular formation of adenosine. We hypothesize that metformin, given at reperfusion, can limit myocardial infarct size due to increased adenosine receptor stimulation. Isolated perfused hearts from Sprague-Dawley rats were subjected to 35 minutes of regional ischemia and 120 minutes of reperfusion. Perfusion with metformin (50 microM) for the first 15 minutes of reperfusion reduced infarct size (percent area at risk) from 42% +/- 2% to 19% +/- 4% (n >or= 6; P < 0.01), which was blocked by a concomitant perfusion with the adenosine receptor antagonist 8-p-sulfophenyltheophylline (100 microM; 43% +/- 3%) or nitrobenzylthioinosine (a blocker of transmembranous adenosine transport; 1 microM; 45% +/- 5%). In addition, intravenous administration of metformin (5 mg/kg) reduced infarct size in a rat in situ model of myocardial infarction (34% +/- 6% vs. 62% +/- 5%; P < 0.01), which was completely abolished by 8-p-sulfophenyltheophylline (61% +/- 3%). We conclude that metformin, given at reperfusion, reduces infarct size in a rat model of myocardial infarction, which is critically dependent on adenosine receptor stimulation, probably via increased intracellular formation of adenosine.