Molecular dynamics (MD) and enhanced sampling simulations are attractive methodologies for the study of biochemical processes due to their ability to describe the structural and dynamic behaviour of biomolecules at atomic resolution.[1] The starting point of these simulations is generally a crystal structure, which does not supply information on the hydrogen atoms and thus raises a challenge on the adequate selection of protonation states in titratable residues. Several methods and utilities have risen to resolve such ambiguity, yet they are far from infallible and therefore it is encouraged to manually revise the suggested states within the system’s region of interest.[2] Nonetheless, in this work we aim to put the spotlight outside that area of interest. The enzyme-inhibitor system trypsin-benzamidine[3] is utilized to evaluate how the protonation states ofHis57,a residue located over 13 Å away from the binding site, can critically influence the computational characterization of protein-ligand binding.Performing spontaneous MD simulations foreach of the three possible protonation states, the number of encountered binding events was observed very different between the two neutral forms and the protonated one of His57. Further assessment with dimensionality reduction techniques demonstrated that the benzamidine binding pathway is dependent on the selected protonation state. A Constant-pH MD approach was additionally carried out, which supplied a more complete picture by accounting for the full equilibrium of the three forms of His57. The present study raises attention to the assignment of protonation states in distal residues in MD simulations, as they can be essential for the proper computational modelling of protein-ligand binding. [1] A. Romero-Rivera, M. Garcia-Borràs, S. Osuna,Chem. Commun.,2017,53, 284-297