Metal contamination of soils is a widespread problem at many current and former industrial and military sites. Lead, As, Cr, and Cd are of particular concern because of their toxicity and potentially harmful effects on the environment. In situ immobilization is a desirable strategy for reducing metal bioavailability through precipitation or adsorption by adding chemical amendments to contaminated soils. Speciation is the key factor in controlling mobility and bioavailability of metals in soils, and knowledge of the mineralogy and geochemistry of contaminant metals is critical for the development of effective remediation strategies. We sampled a Histosol in a peat bog that receives runoff and seepage water from the site of a former lead smelter. The water table is at or slightly above the soil surface for most of the year and the site is covered with common reed (Phragmites australis) and cattail (Typha sp.). There is an intense redox gradient with depth that probably varies as the water table fluctuates. Distinct layers were observed from the surface to a depth of 65 cm. We collected soil samples from different depths during both wet and dry seasons. Along with major elements such as Fe and S, the soil contains 1,700 mg kg^-1 Pb, 2,200 mg kg^-1 As, 930 mg kg^-1 Cr, 210 mg kg^-1 Cd. We used a binocular microscope to select aggregates ~100 to 200 mm in diameter from different depths that had distinctive morphologies and appeared to be inorganic. We used the synchrotron x-ray microprobe on beamline X26A at the National Synchrotron Light Source at Brookhaven National Laboratory to obtain micro x-ray diffraction patterns and micro x-ray fluorescence patterns of each aggregate. Results show a clear pattern of metal speciation changes with depth. The oxidized yellow surface layer was dominant by goethite (α-FeOOH) and a poorly crystalline phase that is probably schwertmannite. Pb and As were highly associated with these Fe oxyhydroxides probably by formation of inner-sphere surface complexes. Gypsum (CaSO₄·2H₂O) is abundant in this layer as well, particularly for samples collected during dry periods. Fe(II)-containing minerals, such magnetite (Fe₃O₄), siderite (FeCO₃), and possibly wustite (FeO) were identified in subsurface layers. These phases may have a strong influence on the fate and transport of chromate due to their highly reactive surfaces and reduction potential. A number of sulfide minerals have been identified in the most reduced horizons at depths >30 cm. They include realgar (AsS; figure below), greigite (Fe²⁺Fe³⁺₂S₄), galena (PbS), sphalerite (Zn, Fe²⁺)S, alacranite (As₄S₄), pyrrhotite (Fe_1-xS), and others. Most of these minerals occur as almost pure phases in the submillimeter aggregates and appear to be secondary phases that have precipitated from solution. Mineralogical and chemical heterogeneity and the presence of phases stable under different redox conditions make this a challenging soil for in situ remediation.