Abstract Excessive Fe 2+ in coastal aquaculture source water will seriously affect the aquaculture development. This study used manganese sand to investigate the removal potential and mechanism of Fe 2+ in coastal aquaculture source water by column experiments. The pseudo‐first‐order kinetic model could better describe Fe 2+ removal process with R 2 in the range of 0.9451–0.9911. More than 99.7% of Fe 2+ could be removed within 120 min while the removal rate ( k ) was positively affected by low initial concentration of Fe 2+ , high temperature, and low pH. Logistic growth (S‐shaped growth) model could better fit the concentration variation of Fe 2+ in the effluent of the column (R 2 >0.99). The Fe 2 breakthrough curve could be fitted by Bohart‐Adams, Yoon‐Nelson, and Thomas models (R 2 >0.95). Smooth slices with irregular shapes existed on the surface of manganese sand after the reaction while Fe content increased significantly on the surface of manganese sand after the column experiment. Moreover, FeO (OH) was mainly formed on the surface of manganese sand after the reaction. Practitioner Points Fe 2+ in coastal aquaculture source water could be removed by manganese ores. The pseudo‐first‐order kinetic model better described the Fe 2+ removal process. FeO (OH) was mainly formed on the surface of manganese sand after the reaction.