For risk assessment of soil contamination, there is a growing need for an approach based on the real risk for living organisms, hence considering the bioavailable fraction of contaminants. The maximum allowable limit of a contaminant in soil, as well as the target for restoration, could be established in relation to a level of bioavailable amount of contaminants considered safe and estimated by standardised procedures. In this work we propose a combination of single and sequential chemical extractions to assess the mobility and plant availability of potentially toxic elements (PTEs) in polluted agricultural soils under remediation. The proposed approach was tested in two pilot sites of Campania potentially polluted by Cu and Zn, before and one year after phytoremediation treatment with Eucalyptus camedulensis L. and Populus nigra L. Specifically, 1 mol L–1 NH4NO3 (readily soluble and phytoavailable PTEs) and 0.05 mol L–1 EDTA at pH 7 (potentially phyto available PTEs) single extractions and the EU-BCR sequential scheme (presumed main geochemical PTE forms) were applied. Overall, in the study sites, the estimation of PTE bioavailability by standardised chemical extractions appeared to be effective in monitoring the Zn and Cu plant-available fractions as well as the trend and effectiveness of phytoremediation treatment. After one year from planting, a significant reduction in Cu and Zn bioavailable amounts in soil was observed (NH4NO3: from 19 to 60% for Cu, from 43 to 92% for Zn; EDTA: from 4 to 27% for Cu and from 8 to 76% for Zn). Plant uptake was positively related with the bioavailability of PTEs in soil, with a general under estimation of Zn extracted in ammonium nitrate. Both single and sequential extractions suggested a higher mobility of Zn with respect to Cu in soil.