The authors studied rate-related hyperpolarization (RRH) in feline Purkinje fibers by standard microelectrode techniques. Restrictions to ionic diffusion are shown to be less in this preparation than in canine preparations. The maximal hyperpolarization (V(H)) and time constant (τ) for the exponentially decaying hyperpolarization following cessation of rapid drive were used as aggregate indices for the underlying rate of sodium pumping, the coupling ratio, and membrane resistance. Changes in V(H) and τ in response to varying stimulation frequencies and durations of rapid drive, to cooling and addition of ouabain, and the changes in (K(+))o and addition of cesium were used to assess the effects on RRH of sodium loading, pump inhibition, and short-circuiting potassium current, respectively. Steady state hyperpolarization was a linear function of stimulation frequency. Increased sodium loading led to an increase in V(H) but no change in τ. Onset and decay of RRH were symmetrical processes. The Q10's of τ and V(H) were 1.57 and 2.00, respectively. For a time after exposure to ouabain, τ and V(H) were unchanged; later, τ increased and V(H) decreased, both linearly with time. Relative membrane resistance decreased with rapid drive and increased with an exponentially decaying time course following cessation of rapid drive. Cesium immediately increased V(H) and decreased τ, and led to progressive deterioration of the fibers. From the steady state equation for RRH, we show that one need not invoke a change in coupling ratio to account for the observed steady state hyperpolarizations. From the nonsteady state equation, the authors show that the Q10 for V(H) is consistent with the Q10 Q10 for Na+,K+-ATPase.