Effects of diphenylhydantoin (DPH) on amphibian atrial myocardium K were investigated using a method which permits both total tissue K and tension response to be monitored continuously. In normal (nondigitalized) preparations, DPH caused a decrease in average K efflux, a net gain of tissue K, and negativeinotropy at low perfusate K concentrations. However, the DPH-induced gain of tissue K was abolished at high perfusate K concentrations while marked negative inotropy was still observed. It is concluded that a gain of tissue K is not the cause of DPH-induced negative inotropy. When digitalis-induced inotropy was associated with tissue K loss, DPH reversed tissue K loss and positive inotropy and caused a decrease in average K efflux. In the presence of toxic effects of digitalis, DPH reversed the K loss and the contracture, but the loss of developed tension was not reversed by DPH. Transmembrane resting potentials and action potential duration were reduced by digitalis and were returned to or above control levels in the presence of DPH. The present findings are consistent with the hypothesis that the therapeutic effect of DPH in digitalis toxicity is brought about by an inhibition of K efflux. This would tend to minimize the loss of tissue K which results from sodium pump inhibition.