The role of free radicals in myocardial reperfusion injury remains controversial. We have developed a new method using ascorbyl free radical (AFR) as a real-time, quantitative marker of free radical generation during myocardial reperfusion. A total of 35 dogs were studied. Twelve open-chest dogs underwent either 5 minutes (n = 5) or 20 minutes (n = 7) of coronary artery occlusion and 30 minutes of reperfusion. Seven additional animals undergoing 20 minutes of coronary occlusion also received the antioxidant enzymes superoxide dismutase and catalase, beginning 10 minutes before occlusion through the end of reperfusion. Exogenous ascorbate was infused intravenously, and the concentration of AFR in the great cardiac vein was continuously measured by electron paramagnetic resonance spectroscopy. Preocclusion AFR concentration was similar in the three groups. Upon reperfusion, AFR rose significantly in each animal group (P < .05). However, the AFR rise in the 20-minute-occlusion group, 38 +/- 17%, was significantly greater than in the 5-minute-occlusion group, 27 +/- 14% (P < .002). In addition, in the animals that received superoxide dismutase and catalase, the rise in the AFR was markedly attenuated, 13 +/- 6% (P < .002). Two dogs that received ascorbate but did not undergo coronary artery occlusion/reperfusion sequences showed no change in coronary venous AFR signal, indicating the stability of the signal over time. Five dogs received ascorbate while undergoing interventions to alter coronary venous flow: intravenous saline, dobutamine, dipyridamole, and nitroglycerin. Coronary venous AFR changes were minimal despite large coronary flow alterations, indicating that the AFR signal is independent of changes in coronary venous flow.(ABSTRACT TRUNCATED AT 250 WORDS)