Dam reservoirs sometimes need to be fully drawn down, an operation that induces sediment resuspension. In the case of contaminated sediments, such resuspension may promote the remobilization of trace metals from the solid to the dissolved phase, threatening the water quality. The objective of this study is to evaluate the kinetic release and removal processes of trace metals during the resuspension of polluted sediments from a dam reservoir. Total contents of trace elements were analyzed in the surface sediments collected in the dam. Ascorbate and sequential extractions were used to assess the trace metal fractionation. Aliquots of wet sediments were resuspended during a week using a 3-L glass reactor. Three solid/solution ratios (2.7, 5.0, and 9.4 g L-1) were tested. The pH, Eh, dissolved oxygen, and temperature were continuously measured, and water samples were collected through a small pipe inserted into the reactor. Evolution of dissolved organic carbon, alkalinity, major ions, and trace element concentrations were measured in the water over the resuspension experiments. Another resuspension experiment was carried out to determine the quantity of elements that could be adsorbed onto the glass wall of the reactor. High levels of As, Cd, Pb, and Zn were measured in the reservoir sediments, compared to the quality guidelines. The temporal evolutions of the dissolved elements were similar during the resuspension experiments, regardless of the mass of sediment used. Depending on the elements, different trends can be highlighted: (1) Cd, Cr, Cu, Ni, V, and Zn were not remobilized; (2) Al, Mo, As, Ba, and Sr continuously increased during the experiments; and (3) Mn, Co, Fe, and Pb increased at the beginning of the experiments and then were removed from the dissolved phase. Co was likely adsorbed onto Mn oxyhydroxides, whereas Pb appeared to be regulated by the Fe oxyhydroxides. The results from sequential extractions did not allow to relate these groups to specific distributions in the solid phase. The greater risk of water quality degradation for this reservoir would be posed by the remobilization of As, which showed the greatest and the most prolonged release during resuspension, because it could not be removed onto particles due to its speciation. Dissolved As reached a maximum of 30 mu g L-1 at the conclusion of the experiments, a concentration above the drinking water standard but well below the acute toxicity levels reported in the literature for fishes.