The many experimental studies reported demonstrate the complexity of what is termed inactivation, the decrease of current flow through sodium channels at maintained depolarization. Even at the normal resting potential of, say, -70 mV for a frog node of Ranvier, ca. 20% of the channels are closed and inactivated, i.e., incapable of passing current on a sudden depolarization, in contrast to the remaining 80% of closed but resting channels. The term inactivation has thus evolved from bulk current ("macroscopic") phenomena and is applied to channels although its single-channel ("microscopic") basis is not entirely clear and may even vary among preparations. It is conceivable that the macroscopic phenomenon may have more than a single microscopic cause; this point will probably not be settled until a physical description of the conformational states of the channel macromolecule becomes available. At any rate, channel transition into an inactivated closed state can be easily affected by numerous reagents of highly diverse chemical nature and, most likely, different primary sites of action as already suggested by the sidedness of effective application, e.g., iodate and endopeptidases to the inside, polypeptide toxins to the outside. But also the search for a common denominator, a secondary target of all these treatments, has not been very successful as demonstrated by the experiments with group-specific reagents. Since modification of inactivation is often accompanied by shifts in the voltage dependence of gating parameters, a target could be the "voltage sensor" of the channel, charged and/or dipolar components of the channel macromolecule that, by being moved in the electric field, somehow induce gating and whose movement is measured as gating current (e.g, Hille, 1984). The fraction of open channels as a function of membrane potential, F(E), may serve as an indicator. It may be simply shifted (to more negative potentials) as by veratridine (Leibowitz et al., 1987) or flattened (reduction of gating charge?) and shifted (in the positive direction) as by Anemonia sulcata toxin II (Ulbricht and Schmidtmayer, 1981) or chloramine-T (Drews, 1987). On the other hand, the steady-state inactivation curve is shifted to more negative potentials by the toxin (Ulbricht and Schmidtmayer, 1981), but to more positive potentials by chloramine-T (Wang, 1984a; Schmidtmayer, 1985). Obviously, modifiers may affect activation and inactivation quite differently, a result that touches on the question as to what extent inactivation derives its potential dependence from activation.(ABSTRACT TRUNCATED AT 400 WORDS)