This dissertation presents results from a series of experiments involving precipitation of ferrous arsenate, redox reactions between various Fe/As couples, and competitive adsorption. Batch and column experiments were made and interpreted quantitatively using geochemical modeling. A new solubility constant was calculated for symplesite, a ferrous arsenate mineral, and geochemical modeling suggests that some arsenic-impacted groundwaters in Bangladesh are super-saturated with respect to this mineral. Oxidation experiments demonstrated that oxygenation of Fe(II) is much faster in the presence of inorganic buffers than when non-complexing organic buffers are used. Fe(II) oxidation was largely unaffected by the presence of the hydroxyl radical scavenger propanol. These findings call into question the classic formulation of Fe(II) oxygenation, and long-accepted kinetic rate constants. During the oxygenation of Fe(II), As(III) competes with Fe(II) for reactive oxidizing species. As(III) oxidation is reduced in the presence of inorganic ligands, most likely because these ligands increase the reactivity of dissolved Fe(II). Competitive adsorption experiments using goethite demonstrated relatively minor competitive effects between As(III) and As(V), and between As(III) and Fe(II). Batch experiments showed that much more Fe(II) was removed from solution than As(III) or As(V) after contacting the goethite surface. This could be explained by the existence of sites which can adsorb Fe(II) but not As. However, surface complexation modeling with this approach could not capture some of the aspects of multi-component adsorption. An alternate explanation could be that upon adsorption Fe(II) transfers an electron into the bulk surface of the goethite, regenerating Fe(III) at the surface and allowing more adsorption to take place. Column experiments were performed to simulate in situ removal of arsenic and iron, and demonstrated that an alternating push-pull configuration can lead to consistent retardation of both solutes. A 'ripening' effect, whereby the in situ process becomes increasingly efficient as more Fe(III) is emplaced on sediment surfaces, was observed at pH 8, where the process increased the amount of iron oxide in the column by more than 50%, even though the column iron oxide concentration was lower than in naturally arsenic-impacted aquifers such as Bangladesh, implying that in situ treatment in such settings is feasible. Advisors/Committee Members: Johnston, Richard, Singer, Philip, Miller, Cass T., Benninger, Larry, Glaze, William, Vasudevan, Dharni, van Gunten, Urs.