In Canada, an estimated land area of 335 hectares is made waste annually for the disposal of neutralization sludge, a by-product of the mining activity. Neutralization sludge is produced following a lime treatment of acidic mining effluents to neutralize the acidity and precipitate the soluble metals. The sludge matrix consists of calcite, gypsum and a large hydrated Fe amorphous phase (ferrihydrite-like) with which several species of potentially-toxic metal contaminants are associated. The long-term environmental impacts of the sludge on surrounding areas is uncertain because the long-term stability of the metals in the sludge is not known. Knowledge of metal speciation would help in assessing the long-term stability of the sludge contaminants and provide a better understanding of how changes in chemical conditions and disposal scenarios may affect changes in sludge chemistry, thereby influencing metal mobility and toxicity. For example, reducing conditions that could potentially develop in a sludge stored under a water cover could favor the reduction and solubilization of Fe(III) to Fe(II), and lead to the mobilization of any contaminant associated with the Fe-phase. In the case of arsenic, the reduction of As(V) to As(III) would further result in the conversion of this contaminant into a more toxic and mobile form than the oxidized As(V) species. An incubation study was initiated to determine the impact of reducing conditions on the dissolution of arsenic in an As-rich neutralization sludge stored under a water cover. The total As concentration in the sludge was 250 mg kg^-1. In the laboratory, the sludge was subjected to water cover treatments in air-tight, plexiglass columns. Three different gas treatments were imposed on the water cover to induce different reduction levels in the sludge (100%N₂, 100%N₂+glucose, 95%N₂:5%H₂). These treatments were compared with a control of oxidized sludge that was saturated with water but with no water cover. The pH, Eh and dissolved metal concentrations were monitored by collecting subsamples of overlying and porewaters at designed intervals (1, 4, 8, 16, 59, 63, 66, 70, 80, 93, 136, 191 and 216 days). After 9 months of reduction, the changes in As and Mn oxidation states in the sludge were characterized using X-ray absorption near-edge structure (XANES) spectroscopy. Electron micro-probe analysis (EPMA) was used to characterize the spatial elemental distribution of As, Fe and Mn in the control and treated sludge samples. The characterization of the sludge was then refined using spatially resolved X-ray fluorescence (XRF) and µ-XANES spectroscopy. In all treatments and throughout the reduction experiment, the dissolved arsenic concentration remained < 5 µg L^-1. Dissolved Mn concentration in the N₂+glucose treatment increased significantly compared to other treatments, indicating reduction activity in this treatment. In agreement with the solution chemistry, Mn and As K-XANES bulk analyses showed that Mn was the redox-active element in the system, while As was very stable. No change in As oxidation state was detected in the reduced samples, and As(V) was still the dominant species in all water-covered sludges at the end of the 9-month reduction. In contrast, the Mn(IV) characterizing the original and control sludge was partly reduced into Mn(II) in the sludge under water. The effect was most pronounced in the N₂+glucose treatment. Results from EPMA suggested that As would be mainly associated with Fe in four different phases (decreasing order of abundance): low-density particles (sludge), high-density particles (iron oxide), some rare grains of pyrite and arseno-pyrite. Mn occurred mainly as Mn-oxide localized micro-spheres, and no As appeared associated with these spheres. The number of Mn micro-spheres decreased after reduction. In the reduced samples, preliminary analysis with spatially resolved XRF and µ-XANES indicated that Mn(II) was preferentially associated with high As concentration. The overall results suggested that Mn acted as a redox barrier to protect As from reduction in the system.