The effects of five freeze-thaw cycles on the dynamic change of dissolved iron in three typical wetland soils (humus marsh soil in Carex lasiocarpa community, meadow marsh soil in Cares meyeriarna community, and meadow albic soil in Calamagrostis angustifolia community)of Sanjiang Plain, Northeast China, was analyzed through in-situ soil column simulation. One freeze-thaw cycle was conducted as freezing at -10 degrees C for 1 d and then thawing at 5 degrees C for 7 d. The thermostatically incubated soils at 5 degrees C were controls. The results showed that most pH and Eh values increased after the first freeze-thaw cycle, and then decreased after the subsequent cycles. 84.4% of the pH values of freeze-thaw treated soils were smaller than that of control, while 82.2% of the Eh values of freeze-thaw treated soils were greater than that of control. Most of the dissolved iron in all soil solutions were Fe3+ ions and colloids, and the reduction of these Fe3+ species were inhibited. The concentrations of Fe2, Fe3+, and total dissolved iron (TFe) of the freeze-thaw treated soils were all smaller than that of controls, with the means of (0.62 +/- 0.08) mg x L(-1) and (1.25 +/- 0.16) mg x L(-1), respectively. The variation trends of pH, Eh, and dissolved iron in the humus marsh soil were significantly different from that in the meadow albic soil. The trends in the meadow marsh soil, as the transitional soil type, were more similar to the meadow albic soil for pH, while more similar to the humus marsh soil for Eh and dissolved iron. Among the three soils, the difference between freeze-thaw treated columns and controls of the second layer were all smaller than that of the third and fourth layer, which indicated that the effect of freeze-thaw cycles were more significant for the upper annular wetland soil layers than the lower layers.