Administration of inhibitors of the Na+/H+ exchanger (NHE) have been shown to produce cardioprotective effects in a number of animal models of ischemia-reperfusion injury; however, controversy still exists as to the efficacy of these agents when administered just before reperfusion. To address this question, the efficacy of several doses of a new selective NHE-1 isoform inhibitor (IC50 for inhibition of 22Na uptake in NHE-1 expressing mouse fibroblast cells = 10.4 +/- 1.0 nM), EMD 85131 (2-methyl-5-methylsulfonyl-1-(1-pyrrollyl)-benzoyl-guanidine), was tested in a canine infarct model in which the left anterior descending coronary artery was occluded for 60 min followed by 3 hr of reperfusion. EMD 85131 (0.75 or 3.0 mg/kg) was infused for 15 min before left anterior descending occlusion or 15 min before reperfusion. Infarct size was determined by use of the triphenyltetrazolium chloride histochemical stain and was expressed as a percent of the area at risk. EMD 85131 (0.75 or 3.0 mg/kg) administered before left anterior descending occlusion produced a marked (*P < .05) and dose-related reduction in IS/AAR (24.3 +/- 3.6, control; 9.3 +/- 3.4%, EMD 0.75; 6.4 +/- 2.3%, EMD 3.0). These two doses of EMD also produced significant (*P < .05) reductions in infarct size/area at risk (12.2 +/- 2.1%, EMD 0.75; 13.0 +/- 2.9%, EMD 3.0) when administered 15 min before reperfusion. These results suggest that selective NHE-1 inhibitors are able to markedly reduce infarct size when given before or during ischemia and also suggest that these compounds may have clinical utility when administered after the initiation of an ischemic insult.