In spite of extensive research during the last decade it has not been possible to prove that endogenously generated hydrogen peroxide or any reduced oxygen species reaches sufficient concentration during reperfusion after myocardial ischemia to contribute significantly to irreversible cell injury. In an attempt to further test this hypothesis we subjected isolated perfused rabbit hearts to 30 min regional ischemia followed by reperfusion and supplied hydrogen peroxide in low levels with or without catalase during the first 30 min of reperfusion and thereafter continued the reperfusion for a total of 120 min. Five different groups were studied: controls, and hearts supplied with 2 microM H2O2, 1 microM H2O2, 1 microM H2O2 + catalase (IU/l) or catalase alone in the initial part of the reperfusion. At the end of 120 min reperfusion, area at risk was measured with fluorescent particles and infarct zone size with tetrazolium staining. The results were: in the control group 32 +/- 5.0% of the risk zone infarcted, in the 2 microM H2O2 group 16.3 +/- 5.6% and in the 1 microM H2O2 group 6.9 +/- 0.8% (P < 0.05 compared to control). The reduction in infarct size was not present when catalase was added to the hydrogen peroxide-containing solution (26.4 +/- 4.5) or if catalase was present alone (22.9 +/- 1.8% infarction). In conclusion, hydrogen peroxide, 1 microM, protected the heart during reperfusion and reduced the amount of cell death after 120 min of reperfusion. The study demonstrated reduction or delay in infarction based only on treatment in the reperfusion period. The mechanism behind this protection remains to be determined.