Repetitive impulse activity in somatic axons results in a progressive reduction in axonal conduction velocity. Since impulse activity produces a concomitant Na-K-ATPase dependent hyperpolarization, changes in membrane potential have hitherto been deemed causal for changes in axonal conduction velocity with depolarization being held to speed conduction and hyperpolarization to retard it. To test this putative causality changes in axonal conduction velocity produced by repetitive activity were examined in single unmyelinated axons innervating the rat cranial meninges. Na-K-ATPase blockade with Ouabain (10m-1mM), by cooling (24–35 °C), by reducing available ATP with cyanide (100–500mM) or by reducing enzymatic substrate (either by lowering the extracellular K⁺ concentration or by replacing extracellular Na⁺ with Li⁺) all increased the magnitude of activity-induced conduction velocity slowing thereby demonstrating that electrogenic Na-K-ATPase activity is not causally responsible for conduction velocity slowing. We propose instead that the most prominent influence of membrane potential on conduction velocity is secondary to its influence on the slow inactivated state of voltage-activated sodium channels. In accord with this idea, the sodium channel blockers lidocaine, carbamazepine and phenytoin (10–500mM) affected activity-induced conduction velocity slowing in a dose dependent manner. At low doses (10–50mM) activity-induced slowing was slightly increased. At higher doses (>100mM) the degree of slowing decreased with increasing dose and at sufficiently high doses, activity-induced slowing was completely blocked. Activity-induced conduction velocity slowing is therefore proposed to be due to a progressive accumulation of voltage-activated sodium channels in their slow inactivated state.