To study the kinetic and steady-state properties of voltage-dependent sodium conductance activation, squid giant axons were perfused internally with either pronase or N-bromoacetamide and voltage clamped. Parameters of activation, tau m and gNa(V), and deactivation, tau Na, were measured and compared with those obtained from control axons under the assumption that gNa oc m3h of the Hodgkin-Huxley scheme. tau m(V) values obtained from the turn-on of INa agree well with control axons and previous determinations by others. tau Na(V) values derived from Na tail currents were also unchanged by pronase treatment and matched fairly well previously published values. tau m(V) obtained from 3 x tau Na(V) were much larger than tau m(V) obtained from INa turn-on at the same potentials, resulting in a discontinuous distribution. Steady-state In (gNa/gNa max - gNa) vs. voltage was not linear and had a limiting logarithmic slope of 5.3 mV/e-fold gNa. Voltage step procedures that induce a second turn-on of INa during various stages of the deactivation (Na tail current) process reveal quasiexponential activation at early stages that becomes increasingly sigmoid as deactivation progresses. For moderate depolarizations, primary and secondary activation kinetics are superimposable. These data suggest that, although m3 can describe the shape of INa turn-on, it cannot quantitatively account for the kinetics of gNa after repolarization. Kinetic schemes for gNa in which substantial deactivation occurs by a unique pathway between conducting and resting states are shown to be unlikely. It appears that the rate-limiting step in linear kinetic models of activation may be between a terminal conducting state and the adjacent nonconducting intermediate.