A lthough a critical factor causing lethal ischemic ventricular arrhythmias, net cellular K loss during myocardial ischemia and hypoxia is poorly understood. We investigated whether selective activation of ATP-sensitive K (KATP) channels causes net cellular K loss by examining the effects of the KATP channel agonist cromakalim on unidirectional K efflux, total tissue K content, and action potential duration (APD) in isolated arterially perfused rabbit interventricular septa. Despite increasing unidirectional K efflux and shortening APD to a comparable degree as hypoxia, cromakalim failed to induce net tissue K loss, ruling out activation of KATP channels as the primary cause of hypoxic K loss. Next, we evaluated a novel hypothesis about the mechanism of hypoxic K loss, namely that net K loss is a passive reflection of intracellular Na gain during hypoxia or ischemia. When the major pathways promoting Na influx were inhibited, net K loss during hypoxia was almost completely eliminated. These findings show that altered Na fluxes are the primary cause of net K loss during hypoxia, and presumably also in ischemia. Given its previously defined role during hypoxia and ischemia in promoting intracellular Ca overload and reperfusion injury, this newly defined role of intracellular Na accumulation as a primary cause of cellular K loss identifies it as a central pathogenetic factor in these settings.