Vagal nerve stimulation (VNS) started prior to, or during, ischemia has been shown to reduce infarct size. Here, we investigated the effect of VNS when started just prior to, and continued during early, reperfusion on infarct size and no-reflow and studied the underlying mechanisms. For this purpose, swine (13 VNS, 10 sham) underwent 45 min mid-LAD occlusion followed by 120 min of reperfusion. VNS was started 5 min prior to reperfusion and continued until 15 min of reperfusion. Area at risk, area of no-reflow (% of infarct area) and infarct size (% of area at risk), circulating cytokines, and regional myocardial leukocyte influx were assessed after 120 min of reperfusion. VNS significantly reduced infarct size from 67 ± 2 % in sham to 54 ± 5 % and area of no-reflow from 54 ± 6 % in sham to 32 ± 6 %. These effects were accompanied by reductions in neutrophil (~40 %) and macrophage (~60 %) infiltration in the infarct area (all p < 0.05), whereas systemic circulating plasma levels of TNFα and IL6 were not affected. The degree of cardioprotection could not be explained by the VNS-induced bradycardia or the VNS-induced decrease in the double product of heart rate and left ventricular systolic pressure. In the presence of NO-synthase inhibitor LNNA, VNS no longer attenuated infarct size and area of no-reflow, which was paralleled by similarly unaffected regional leukocyte infiltration. In conclusion, VNS is a promising novel adjunctive therapy that limits reperfusion injury in a large animal model of acute myocardial infarction.