1. Intracellular Na activity, aiNa, was measured in voltage-clamped sheep cardiac Purkinke fibres. 2. Increasing [Rb]0 from 0 to 4 mM in K-free solutions (at a fixed membrane potential) decreased aiNa. Further increases of [Rb]0 (up to 20 mM) had little or no effect. 3. Following exposure to Rb-free, K-free solution, the addition of a test concentration of Rb produced an exponential decrease of aiNa. The rate constant of decay of aiNa increased with increasing [Rb]0 over the measured range (0-20 mM). 4. The accompanying electrogenic Na pump current transient decayed with the same rate constant as aiNa over the range of [Rb]0 examined. During this decay the electrogenic Na pump current was a linear function of aiNa. Increasing [Rb]0 increased the steepness of the dependence of the electrogenic current on aiNa. 5. A constant fraction of the net Na efflux was electrogenic. This fraction was not affected by varying [Rb]0 over the range 0-20 mM. 6. Using a simple model, it is shown that the dependence of steady-state aiNa on [Rb]0 is half-saturated by less than 1 mM-[Rb]0. The rate constant of decay of aiNa and the slope of the relationship between electrogenic Na pump current and aiNa are, however, better fitted with a lower affinity for Rb (K0.5 = 4 mM-[Rb]0). 7. Depolarizing the membrane potential with the voltage clamp decreased aiNa; hyperpolarization increased it. These effects persisted in the presence of 10(-5) M-strophanthidin. An effect of membrane potential on the net passive Na influx can account for the observations. 8. The effects of membrane potential on the net passive Na influx were examined by measuring the maximum rate of rise of aiNa at different holding potentials after inhibiting the Na-K pump in a K-free, Rb-free solution. Depolarization decreased the Na influx. 9. Using the constant field equation, the net passive Na influx was used to estimate the apparent Na permeability coefficient, PNa. This was between 0.8 x 10(-8) and 1.5 x 10(-8) cm sec-1.