Isolated rabbit left atrial preparations were perfused with Tyrode's solutions containing 1 to 10 µg/ml (4 x 10^-6-4 x 10^-5M) diphenylhydantoin (DPH), 2.6-5.6 mM K⁺, and 154-308 mM Na⁺. Steady-state transmembrane resting and action potentials were recorded from these preparations with glass microelectrodes at stimulation rates ranging from 0.2 to 3/sec. DPH had little or no effect on the relationship between extracellular [K⁺] and membrane resting potential. Action potential overshoot was generally decreased by 5 and 10 µg/ml DPH and increased by 1 µg/ml DPH at stimulation rates of 2 and 3/sec in the presence of increased [K⁺]. DPH and increased [K⁺] acted synergistically to shorten action potential duration (measured at 50% repolarization). The effect of DPH on phase 0 of the action potential (measured as action potential rise time between 10 and 50% and 50 and 90% depolarization) was markedly dependent upon drug concentration, extracellular [K⁺] and stimulation rate. The lowest concentration of DPH (1 µg/ml) usually shortened action potential rise time, particularly when it had been prolonged by increasing extracellular [K⁺]. Conversely, the highest concentration of DPH (10 µg/ml) and increased [K⁺] acted synergistically to prolong action potential rise time (i.e., decrease depolarization rate). When present, the depressant effect of DPH on membrane depolarization was rapidly antagonized by increasing extracellular [Na⁺]. It is proposed that DPH may either enhance or depress (like quinidine) membrane activity in atrial tissue, and that both the direction and magnitude of effect are strongly dependent upon drug concentration, ionic milieu, and heart rate.