Scorpion b-toxins are peptides of 55–70 residues; they bind specifically to voltage-gated sodium channels (NaV channels) and cause use-dependent subthreshold channel openings. A voltage sensor trapping model was proposed in which b-toxins preactivate NaV channels by trapping the voltage sensor of channel domain-2 in an activated, outward position upon depolarization of the membrane (Cestèle et al. 1998, Neuron 21:919). These preactivated channels require less activation energy in subsequent depolarization cycles and therefore open at potentials below the normal activation threshold. However, this excitatory action of b-toxins is often overlaid by a not yet understood depressant effect, i.e. b-toxins tend to simultaneously inhibit NaV channels indicating that the mechanism of gating modification is more complex. We analyzed the effects of b-toxin Tz1 from Tityus zulianus on heterologously expressed NaV1.4 channels using the whole-cell patch-clamp method. A quantitative data analysis revealed three parameters characterizing the effects of Tz1: the toxin-induced shift dV in activation threshold of the channel fraction Ps and the channel fraction inhibited by the toxin, Pi. At a holding potential of -160 mV and a stimulation rate of 0.1 Hz, 5 muM Tz1 inhibited most NaV1.4 channels (Ps=0.100.02, Pi=0.550.02, dV=-34.60.9 mV; n=5) indicating its depressant function under this condition. In contrast, an increase of the stimulation rate to 2 Hz facilitated the excitatory function of Tz1 as indicated by an increase in Ps to 0.480.05 and a drop in Pi to 0.170.06 (n=5). Notably, the fraction of all Tz1-modified channels, i.e. the sum of Ps and Pi, was identical under both conditions. This indicates that the use-dependent change in the mode distribution does not involve unbinding and rebinding of the toxin. The voltage-dependence of the change in steady-state mode distribution was characterized by an apparent gating charge of one electron charge. The structural involvement of the voltage-sensor in channel domain-2 was assessed by mutation of individual positively charged residues in S4. Neutralization of R663, the outermost charge in the voltage sensor, facilitated the depressant mode of Tz1. This demonstrates the importance of R663 for the outward movement of the sensor and consequently for channel opening. In contrast, neutralization of R669 in the center of the sensor enhanced the excitatory mode of the toxin. Therefore, R669 is involved in the inward movement of an activated domain-2 sensor. In conclusion, our data support a model in which Tz1 is capable of stabilizing two conformations of the voltage sensor in domain-2 of NaV1.4; a preactivated, outward position leading to channels that open at subthreshold potentials and a deactivated, inward position preventing channels from opening. In addition, analysis of b-toxin action reveals important information on the voltage-sensor movement during channel gating.