A series of 9 soil samples were taken at a timber treatment site in SW France where Cu sulphate and chromated copper arsenate (CCA) have been used as wood preservatives (Sites P1 to P9) and one soil sample was collected at an adjacent site on the same soil type (Site P10). Copper was a major contaminant in all topsoils, varying from 65 (Soil P5) to 2600 mg Cu kg−1 (Soil P7), exceeding background values for French sandy soils. As and Cr did not accumulate in soil, except at Site P8 (52 mg As kg−1 and 87 mg Cr kg−1) where CCA-treated posts were stacked. Soil ecotoxicity was assessed with bioassays using radish, lettuce, slug Arion rufus L., and earthworm Dendrobaena octaedra (Savigny). There were significantly differences in lettuce germination rate, lettuce leaf yield, radish root and leaf yields, slug herbivory, and earthworm avoidance. An additional bioassay showed higher negative impacts on bean shoot and root yields, Rhizobium nodule counts on Bean roots, and guaiacol peroxidase activity in primary Bean leaves for soil from Site P7, with and without fertilisation, than for soil from Site P10, despite both soils having a similar value for computed free ion Cu2+ activity in the soil solution (pCu2+). Beans grown in soil from Site P7 that had been fertilised showed elevated foliar Cu content and phytotoxic symptoms. Soils from Sites P7 (treatment plant) and P6 (storage of treated utility poles) had the highest ecotoxicity, whereas soil from Site P10 (high organic matter content and cation exchange capacity) had the lowest. Except at Site P10, the soil factor pCu2+ computed with soil pH and total soil Cu could be used to predict soil ecotoxicity.< Réduire