Lidocaine is a Na+ channel blocker that is highly effective for the treatment of ventricular tachyarrhythmias, but is largely ineffective against atrial arrhythmias. If is not known if this differential efficacy is the result of differences in lidocaine inhibition of atrial v ventricular Na+ channels. The purpose of the present study was to characterize lidocaine block of Na+ channels in human atrium and ventricle. We used the whole cell voltage clamp technique with low external and internal Na+ concentrations (5 mM) to study the Na+ current (INa) in single human atrial and ventricular cells isolated enzymatically from specimens obtained during surgery. We found that tonic block of peak INa by lidocaine (200 microM, holding potential = -140 mV, 0.1 Hz, at 17 degrees C) was not voltage dependent in either cell type. Reduction of maximal peak Na+ conductance in 41 atrial cells (19.8 +/- 2.7%) and nine ventricular cells (22.6 +/- 1.7%) was virtually identical. The rate of onset of block development was determined during depolarization to either -80 mV or -20 mV. The time course of onset of block was described by a single exponential at -80 mV and by a double exponential at -20 mV. When the rate of block onset during a single conditioning depolarization was compared to that which developed during conditioning by a train of brief pulses (3 ms, 30 Hz), onset was faster during the pulse train. The results were nearly identical for atrial and ventricular INa. The time constants of recovery from block following either single pulse or multiple-pulse conditioning did not differ. These data suggest that lidocaine binds to both the activated and inactivated states of the human cardiac Na+ channel. Using an analytical method based upon the Guarded Receptor Hypothesis, we calculated apparent rate constants describing lidocaine's interaction with the three primary states of the human Na+ channel (resting, activated and inactivated). Rate constants were similar to those reported for other mammalian species. Our results demonstrate that lidocaine block of INa is virtually identical for human atrial and ventricular cells; thus additional mechanisms must be invoked to explain the differential efficacy of lidocaine against ventricular as compared to atrial dysrhythmias.