Gating current (Ig) underlying Na-channel activation is large enough to enable resolution of components both preceding and paralleling Na conductance (gNa) turn-on. For large depolarizations (beyond +20 mV), an additional "slow phase" of Ig is observed during a time when Na activation is already complete, but when K-channel opening is just becoming detectable. If Na- and K-channel gating are similar, the slow kinetics and long delay for K activation predict that K channel Ig must be relatively small and slow. Externally applied dibucaine almost totally blocks gNa and greatly reduces the fast (Na channel) Ig without altering gK or the Ig slow phase. The slow phase of Ig depends in part of the presence of functional K channels. Selective diminution in amplitude of the slow phase is consistently observed after a 30-min perfusion with both external and internal K-free media, a procedure which destroys nearly all K channels. This decrease of Ig amounts to approximately 10% of the total charge movements at +40 to +80 mV, with gating charge and K channels disappearing in a ratio of less than 1 e- per picosiemens of gK. These findings are consistent with the idea that part of the Ig slow phase represents gating current generated by the early steps in K-channel activation.