The performances of five Diffusive Gradients in Thin Films (DGT) binding gels, namely 3-mercapto-functionalized silica (3MP), ferrihydrite (Fh), Metsorb®, zinc ferrite (ZnFe2O4), and Zirconium oxide (ZrO2), were evaluated for in situ determination of As speciation in water and sediments. A combination of batch experiments at various pH (without addition of buffers) and in the presence of reduced species (Mn2+, Fe2+ and HS-),time-series experiments in oxic waters, and in situ deployment in anoxic river sediments has permitted to evaluate the potential interferences among the binding gels. Firstly, the efficiency of each DGT binding gel dedicated to total As (i.e., Fh, Metsorb®, ZnFe2O4 and ZrO2) or As(III) (i.e., 3MP) determination confirms that the determination of As species is possible in oxic freshwater and seawater over 96 h for a wide range of pH (5−9). Secondly, concerning the deployment in river sediment, high HCO3- concentrations have a little negative effect only on the DGT performances of the iron(III)-binding gels (i.e, Fh and ZnFe2O4). Thirdly, the presence of sulfides does not show any effect on the DGT uptake of As, but strongly affects the elution factor parameter. Discrepancies in elution between the different binding gels potentially result in precipitation of orpiment, especially in 1 mol L−1 HNO3. A correction of the classical elution factor derived from batch experiments was applied to provide more representative results. Finally, this study shows the difficulties to determine As speciation in anoxic sediments, and suggests that corrections of the elution factor may be required as a function of the species present in the deployment matrices.