In this study the importance of sea salt aerosol (SSA) size representation in a global transport model is investigated. For this purpose the Navy Aerosol Analysis and Prediction System (NAAPS) model is employed in a number of SSA simulations. A new dry deposition velocity parameterization is implemented into NAAPS in order to more physically represent deposition processes in the model. SSA size distribution is divided into size bins using two different partition procedures: the previously used iso-log method and the iso-gradient method, which relies on size-dependence of deposition processes. The global SSA simulations are analyzed in terms of the total sea salt mass and the average SSA optical thickness. The results indicate that there is a large dependence of the total mass and average aerosol optical depth on the number of size bins used to represent the aerosol size distribution. The total SSA mass is underestimated by 20% if 2 instead of 15 (reference) size intervals are used. The average aerosol optical depth underestimation is even higher and reaches over 35%. Such large differences can have substantial implications on the accuracy of SSA radiative forcing simulations in climate models. A comparison of the two division procedures shows that the simulations with the iso-gradient intervals are more accurate than the iso-log ones if at least 6 size bins are used. This result indicates that the more physically based division scheme can offer better performance and reduce computational cost of global aerosol transport models.