Three mesocosm wetlands (250 cm × 100 cm × 100 cm) with different wetland plants (Calamgrostis angustifolia, CA, Carex lasiocarpa, CL, and C. angustifolia/C. lasiocarpa mixture, AL, respectively) and hydrologic regimes were set to test migration and retention of exogenous dissolved iron ((NH₄)₂Fe(SO₄)₂of 40 mg Fe(II) L^？1) in the Sanjiang Plain Wetland in northeast China. The experiment was designed as two stages: open migration period (OMP) for 1.5 d and close retention period (CRP) for 28.5 d. Based on the outflow Fe(II) concentration during the OMP, retention efficiencies (RE) and iron retention fluxes adjusted by area (RF_ad) in the three mesocosm wetlands were calculated, and the migration of iron were modeled using the first-order kinetic model. Outflow pH decreased gradually from a weak alkaline condition to a weak acid condition during the OMP, and then increased during the CRP, while outflow Eh and DO decreased during the experiment. The three mesocosm wetlands had considerable RE ranging from 75% to 98%, with the averaged RF_ad of 4.31 ± 0.17, 4.20 ± 0.16, and 4.37 ± 0.13 g m^？2 h^？1 for CA, CL, and AL, respectively. The reduction conditions in the mesocosm wetlands developed after 4 d or 12 d and the former retained iron during the OMP became mobile and discharged primarily in the form of Fe(III). The first-order kinetic model could simulate the outflow concentration of dissolved iron during the OMP (R² = 0.91, 0.69, and 0.68 for CA, CL, and AL, respectively), while the outflow dissolved iron during the CMP was difficult to model because the changed pH and Eh conditions in the mesocosm wetlands cause the former precipitated iron to be mobile after several days.