In crystallographic structures of NavAb and other bacterial sodium channels the P-loop domain contains P2 helices, which are absent in potassium channels. Here we asked whether these structures can advance understanding of experimental data accumulated for eukaryotic sodium channels. Using published experimental data on interactions of mu-conotoxins with Nav1.X channels, we docked GIIIA, PIIIA and KIIIA in the NavAb-based model of Nav1.4. Docking of GIIIA was based on specific channel-toxin contacts described in published experimental studies. Importantly, these specific contacts were readily reproduced in our computations with Monte Carlo-energy minimizations without modifications of the template backbone geometry. Computed energies of specific interactions correlated with experimental estimations. Predicted orientation of the GIIIA was used to dock PIIIA and KIIIA. The obtained toxin-channel complexes are consistent with mutational data and voltage-dependence of toxin action. Particularly, tetrodotoxin can pass between Nav1.4 and the channel-bound KIIIA to reach its binding site in the selectivity filter. KIIIA and some GIIIA and PIIIA mutants are known to incompletely block the current. To understand this phenomenon, we Monte Carlo-minimized the energy of toxin-channel complexes from many starting points with randomly placed sodium ions and superimposed low-energy structures to visualize constellations of the ions. Uninterrupted pathways of sodium ions between the extracellular space and the selectivity filter where seen only when at least one outer carboxylate was not salt-bridged to the toxin. A good correlation was found between the modeling results and experimental data on complete and incomplete channel block by the native and mutant toxins. Thus, the NavAb structure advances understanding of permeation and block of eukaryotic sodium channels. Our study suggests similar folding of the outer-pore region in eukaryotic and prokaryotic sodium channels. Supported by RFBR-13-04-00724 to DBT and NSERC to BSZ.