Between 37 and 20 C., the resting potential of sheep ventricular and Purkinje fibers was found to be proportional to the absolute temperature. Prom 20 to 0 C., the potential drop was 3 times greater than that predicted from the theory of a concentration potential. Rapid heating by high-frequency current of previously cooled preparations, revealed no time lag between the rise of potential and the rise of temperature. Fibers kept in a Na-free solution still showed a large temperature sensitivity. With poisoned fibers, the resting potential was almost proportional to the absolute temperature. The results appear to contradict the hypothesis that extracellular K accumulation in the cold is the cause of the potential drop. They also exclude the hypothesis that a relative rise of the Na-inward current is the reason for depolarization. They support the idea that “active” extrusion of Na ions normally contributes to the resting potential, and that the pumping rate decreases considerably in the cold.