Studies in vitro suggested that inflammatory cytokines could cause myocardial dysfunction. However, the detailed mechanism for the cytokine-induced myocardial dysfunction in vivo remains to be examined. We thus examined this point in our new canine model in vivo, in which microspheres with and without IL-1beta were injected into the left main coronary artery. Left ventricular ejection fraction (LVEF) was evaluated by echocardiography for 1 wk. Immediately after the microsphere injection, LVEF decreased to approximately 30% in both groups. While LVEF rapidly normalized in 2 d in the control group, it was markedly impaired in the IL-1beta group even at day 7. Pretreatment with dexamethasone or with aminoguanidine, an inhibitor of inducible nitric oxide synthase, prevented the IL-1beta-induced myocardial dysfunction. Nitrotyrosine concentration, an in vivo marker of the peroxynitrite production by nitric oxide and superoxide anion, was significantly higher in the myocardium of the IL-1beta group than in that of the control group or the group cotreated with dexamethasone or aminoguanidine. There was an inverse linear relationship between myocardial nitrotyrosine concentrations and LVEF. These results indicate that IL-1beta induces sustained myocardial dysfunction in vivo and that nitric oxide produced by inducible nitric oxide synthase and the resultant formation of peroxynitrite are substantially involved in the pathogenesis of the cytokine-induced sustained myocardial dysfunction in vivo.