This paper explores the impact of Mn doping in the battery electrode material Li(Mn)FePO4 by computational methods. Structures are generated at up to 6.25% Li and 3.125% Mn content with varying Li, Fe, and Mn configurations. The oxidation state of the Fe and Mn ions are explicitly set using occupation matrix control, while the structures are optimized using DFT + U. Although Mn is redox active and can exist in both the +III and +II states, there is a strong preference for the latter. We show that the charge-compensating electrons prefer to occupy the Mn sites rather than the Fe sites during Li insertion. This results in Mn(II) and Fe(II) ions that interact electrostatically with the positively charged Li ions. The DFT + U data are consequently used to parametrize a Coulomb potential, incorporating short-range two-body corrections, and utilized within a Monte Carlo approach. The short-range order between the atomic species are then quantified by their coordination numbers at varying temperatures. The Monte Carlo sampling demonstrates lower Li-Li clustering in the Li(Mn)FePO4 system at temperatures below 500 K and thus more disorder as compared to the undoped system. On average, this results in less cluster formation, which correlates with the enlarged solid solution region found in the LiMnFePO4 phase diagram.