There has been considerable controversy regarding the role of oxygen free radicals as important mediators of cell damage in reperfused myocardium. This controversy regards whether superoxide and hydroxyl free radicals are generated on reperfusion and if these radicals actually cause impaired contractile function. In this study, EPR studies using the spin trap 5,5-dimethyl-1-pyroline-n-oxide (DMPO) demonstrate the formation of .OH and R. free radicals in the reperfused heart. EPR signals of DMPO-OH, aN = aH = 14.9 G, and DMPO-R aN = 15.8 G aH = 22.8 G are observed, with peak concentrations during the first minute of reperfusion. It is demonstrated that these radicals are derived from .O2- since reperfusion in the presence of enzymatically active recombinant human superoxide dismutase markedly reduced the formation of these signals while inactive recombinant human superoxide dismutase had no effect. On reperfusion with perfusate pretreated to remove adventitial iron, the concentration of the DMPO-OH signal was increased 2-fold and a 4-fold decrease in the DMPO-R signal was observed demonstrating that iron-mediated Fenton chemistry occurs. Hearts reperfused with recombinant human superoxide dismutase exhibited improved contractile function in parallel with the marked reduction in measured free radicals. In order to determine if the reperfusion free radical burst results in impaired contractile function, simultaneous measurements of free radical generation and contractile function were performed. A direct relationship between free radical generation and subsequent impaired contractile function was observed. These studies suggest that superoxide derived .OH and R. free radicals are generated in the reperfused heart via Fenton chemistry. These radicals appear to be key mediators of myocardial reperfusion injury.