Amagnesia-pullulan (MgOP) composite has been developed to remove phosphate from a synthetic solution. In the present study, the removal of phosphate by MgOP was evaluated in both a batch and dynamic system. The batch experiments investigated the initial pH effect on the phosphate removal efficiency from pH 3 to 12 and the effect of co-existing anions. In addition, the adsorption isotherms, thermodynamics, and kinetics were also investigated. The results from the batch experiments indicate that MgOP has encouraging performance for the adsorption of phosphate, while the initial pH value (3-12) had a negligible influence on the phosphate removal efficiency. Analysis of the adsorption thermodynamics demonstrated that the phosphate removal process was endothermic and spontaneous. Investigations into the dynamics of the phosphate removal process were carried out using a fixed bed of MgOP, and the resulting breakthrough curves were used to describe the column phosphate adsorption process at various bed masses, volumetric flow rates, influent phosphate concentrations, reaction temperatures, and inlet pH values. The results suggest that the adsorption of phosphate on MgOP was improved using an increased bed mass, while the reaction temperature did not significantly affect the performance of the MgOP bed during the phosphate removal process. Furthermore, higher influent phosphate concentrations were beneficial towards increasing the column adsorption capacity for phosphate. Several mathematic models, including the Adams-Bohart, Wolboska, Yoon-Nelson, and Thomas models, were employed to fit the fixed-bed data. In addition, the effluent concentration of magnesium ions was measured and the regeneration of MgOP investigated.