There is an ongoing discussion in the literature if simulations of electrochemical reactions should be performed at constant electrode potential. In this context we examine the timescales at which electrochemical processes take place. We take molecular dynamics simulation of hydrogen desorption from graphene and gold as examples. Charge transfer processes are governed by the accompanying solvent reorganization, and their speed is determined by the solvent dynamics, whose timescale is of the order of picoseconds – the exact value depends on the details of the system. In contrast, double layer relaxation, which controls the electrode potential, requires times of the order of nanoseconds. We suggest that this difference is due to the fact, that solvent reorganization is a local phenomenon, while double layer relaxation involves the whole solvent between working and reference electrodes, which is of macroscopic dimensions. We conclude, that simulations of electrochemical reactions should not be performed at constant potential. In contrast, calculations for thermodynamic properties can be performed at constant potential or at constant charge, since the question of timescales does not arise in this case.