Voltage-gated ion channels are key players in fast neuronal signalling. Detailed knowledge about channel gating is essential for our understanding of channel function in general and of drug action of channels in particular. Despite a number of recent atomic channel structures, the opening of voltage-gated channels is the subject of heated debates. Here we will discuss two of the controversies: one concerning the mechanism of opening and closing the pore, and the other concerning the location and movement of the voltage sensor. The channels were originally suggested to open at a conserved proline rich sequence (PVP) at the intracellular end of the transmembrane segment 6 (S6). The crystallization of a channel in the open state instead suggested an opening involving a conserved glycine hinge located in the middle portion of S6. Based on pharmacological studies, autodocking and molecular dynamics simulations we have found support for the PVP-bend model. The voltage sensor, transmembrane segment 4 (S4), was originally suggested to be buried in the channel protein, undergoing a helical-screw-like motion to open the channel. A recent crystallographic study suggested that S4 is located in the periphery, facing lipid, and undergoing a paddle-like motion to open the channel. We have found experimental evidence for a novel helical-screw model; with the voltage sensor moving in a screw-like fashion but being located in the periphery of the channel. This model opens up for understanding how lipophilic drugs and toxins directly affect the voltage sensor.