Spinal neurons undergo large cyclic membrane potential oscillations during fictive locomotion in lamprey. It was investigated whether these oscillations were due only to synaptically driven excitatory and inhibitory potentials or if voltage-dependent inward conductances also contribute to the depolarizing phase by using N-(2,6-dimethylphenyl carbamoylmethyl)triethylammonium bromide (QX-314) administered intracellularly during fictive locomotion. QX-314 intracellularly blocks inactivating and persistent Na+ channels, and in some neurons, effects on certain other types of channels have been reported. To detail the effects of QX-314 on Na+ and Ca2+ channels, we used dissociated lamprey neurons recorded under whole cell voltage clamp. At low intracellular concentrations of QX-314 (0.2 mM), inactivating Na+ channels were blocked and no effects were exerted on Ca2+ channels (also at 0.5 mM). At 10 mM QX-314, there was, however a marked reduction of I(Ca). In the isolated spinal cord of the lamprey, fictive locomotion was induced by superfusing the spinal cord with Ringer's solution containing N-methyl-D-aspartate (NMDA), while recording the locomotor activity from the ventral roots. Simultaneously, identified spinal neurons were recorded intracellularly, while infusing QX-314 from the microelectrode. Patch electrodes cannot be used in the intact spinal cord, and therefore "sharp" electrodes were used. The amplitude of the oscillations was consistently reduced by 20-25% in motoneurons (P < 0.05) and unidentified spinal neurons (P < 0.005). The onset of the effect started a few minutes after impalement and reached a stable level within 30 min. These effects thus show that QX-314 causes a reduction in the amplitude of membrane potential oscillations during fictive locomotion. We also investigated whether QX-314 could affect glutamate currents by applying short pulses of glutamate from an extracellular pipette. No changes were observed. We also found no evidence for a persistent Na+ current in dissociated neurons, but these cells have a much-reduced dendritic tree. The results indicate that there is an inward conductance, which is sensitive to QX-314, during membrane potential oscillations that "boosts" the synaptic drive during fictive locomotion. Taken together, the results suggest that inactivating Na+ channels contribute to this inward conductance although persistent Na+ channels, if present on dendrites, could possibly also contribute to shaping the membrane potential oscillations.