1. The vaseline-gap voltage-clamp method has been applied to the study of ionic currents in cut pieces from innervated and 5--7 days denervated rat skeletal muscle fibres. 2. Kinetic analysis of sodium currents in innervated rat muscle showed them to be similar to those in frog muscle, except that rat sodium channels are activated at slightly more negative potentials. Peak sodium conductances of 40--50 m-mho/cm2 were measured, corresponding to values of GNa of 100--120 m-mho/cm2. 3. The permeability sequence of the sodium channel to several organic and inorganic cations is Li+ greater than Na+ greater than hydroxylammonium greater than hydrazinium greater than guanidinium approximately ammonium greater than K+. TMA+ and Ca2+ were not measurably permeant. 4. Denervation appears to shift activation and inactivation parameters of sodium currents by approximately 10 mV to more negative potentials, but does not appreciably affect the maximum peak sodium conductance or the time constants for activation and inactivation. 5. Dose--response curves for block by tetrodotoxin in innervated fibres are fitted well by assuming binding of toxin to a single population of channels with a dissociation constant of about 5 nM. In denervated fibres there appears in addition a second population of channels with a dissociation constant in the micromolar range. These relatively toxin-insensitive channels respond less rapidly to potential changes, and can contribute up to 25--30% of the total sodium conductance. 6. The potassium currents of innervated rat muscle were similar to those of frog muscle in their voltage dependence of activation. 7. The time constant for inactivation of the sodium current, tau h, at -13 mV showed a temperature dependence measured between 10 and 20 degrees C equivalent to an average Q10 of 2 . 3. The Q10 for the time constant of activation of the potassium current tau n, averaged 2 . 5 between -40 mV and +40 mV, measured over the same temperature range.