The respective importance of flow and cellular viability in determining initial myocardial thallium uptake was studied in reperfused and nonreperfused experimental myocardial infarction. Open-chest dogs were subjected to permanent coronary artery occlusion of 70-minute (n = 3) or 5-hour duration (n = 5), or to a 3-hour temporary occlusion followed by reflow (n = 14). Thallium uptake 10 minutes after intravenous injection was compared directly with radioactive microspheres in myocardial samples from excised hearts. Triphenyl tetrazolium chloride staining was used to differentiate necrotic and viable samples with confirmation by electron microscopy. In nonreperfused infarcts, thallium uptake occurred despite necrosis, and a close correlation was found between thallium uptake and regional myocardial blood flow. In reperfused infarcts, thallium uptake again occurred, but was reduced relative to flow in necrotic myocardium and, to a lesser extent, in reperfused viable areas. However, because of the high levels of reflow, actual thallium uptake was often more than 50% of normal in reperfused necrotic regions. This study demonstrates that the presence of thallium uptake is an unreliable indicator of myocardial injury and that reperfused necrotic tissue may have remarkably high levels of thallium uptake.