Tetrodotoxin (TTX) is a potent neurotoxin naturally found in terrestrial and marine animals targeting the voltage-gated sodium channels. Historically, TTX has raised food safety concerns mainly in the Asian countries due to the consumption of the typical pufferfish-derived delicacy fugu. However, the diffusion of TTX is getting wider today, reasonably threating in a close future a broader number of consumers than before. The understanding of TTX group toxicity is still incomplete as most of the analogues and metabolites found together with TTX are largely understudied. This prevents the establishment of a solid background for risk assessment and additional data have been claimed to timely foster the assessment of TTX toward a group-based approach. However, the high costs in sourcing TTX analogues make practically unfeasible the wide-scale assessment using experimental trials. The toxicological assessment in silico may succeed in extending data on compounds poorly affordable, hierarchizing compound to focus experiments and supporting the hazard identification. Therefore, the present work investigated the toxicodynamic of TTX, analogues and metabolites by using a molecular modeling approach. In the framework of the hazard identification, the model analyzed TTX analogues never tested before assessing qualitatively their potential toxicity in comparison to TTX. While the analogues from TTX bearing species appeared to be less toxic than TTX, some human metabolites showed a better interaction with the toxin binding site. Such results suggest that human metabolism may partially fail in preventing the interaction with the biological target. Therefore, the identification and assessment of human metabolites should be done to support the decision making process from a more informed perspective.