Major uncertainties exist with respect to the aqueous geochemical evolution of the Martian surface. Considering the prevailing cryogenic climates and the abundance of salts and iron minerals on Mars, any attempt at comprehensive modeling of Martian aqueous chemistry should include iron chemistry and be valid at low temperatures and high solution concentrations. The objectives of this paper were to (1) estimate ferrous iron Pitzer-equation parameters and iron mineral solubility products at low temperatures (from < 0 ??C to 25 ??C), (2) incorporate these parameters and solubility products into the FREZCHEM model, and (3) use the model to simulate the surficial aqueous geochemical evolution of Mars. Ferrous iron Pitzer-equation parameters were derived in this work or taken from the literature. Six new iron minerals [FeCl2??4H2O, FeCl2??6H2O, FeSO4??H2O, FeSO4??7H2O, FeCO3, and Fe(OH)3] were added to the FREZCHEM model bringing the total solid phases to 56. Agreement between model predictions and experimental data are fair to excellent for the ferrous systems: Fe-Cl, Fe-SO4, Fe-HCO3, H-Fe-Cl, and H-Fe-SO4. We quantified a conceptual model for the aqueous geochemical evolution of the Martian surface. The five stages of the conceptual model are: (1) carbonic acid weathering of primary ferromagnesian minerals to form an initial magnesium-iron-bicarbonate-rich solution; (2) evaporation and precipitation of carbonates, including siderite (FeCO3), with evolution of the brine to a concentrated NaCl solution; (3) ferrous/ferric iron oxidation; (4) either evaporation or freezing of the brine to dryness; and (5) surface acidification. What began as a dilute Mg-Fe-HCO3 dominated leachate representing ferromagnesian weathering evolved into an Earth-like seawater composition dominated by NaCl, and finally into a hypersaline Mg-Na-SO4-Cl brine. Weathering appears to have taken place initially under conditions that allowed solution of ferrous iron [low O2(g)], but later caused oxidation of iron [high O2(g)]. Surface acidification and/or sediment burial can account for the minor amounts of Martian surface carbonates. This model rests on a large number of assumptions and is therefore speculative. Nevertheless, the model is consistent with current understanding concerning surficial salts and minerals based on Martian meteorites, Mars lander data, and remotely-sensed spectral analyses. ?? 2003 Elsevier Ltd.