This thesis was executed in the context of the historical metal pollution in the Campine (Kempen) region in north-eastern Belgium. Due to metal refining activities in the past, an extended area of about 700 km² is diffusely polluted with mainly cadmium (Cd) and zinc (Zn). The historical soil pollution is still causing human-toxicological and ecotoxicological risks, through metal leaching to groundwater and accumulation in the food chain. Moreover, the Campine region is characterized by poor sandy soils, aggravating the risks for metal dispersion in the environment. Conventional soil sanitation techniques are technically and financially inadequate to tackle the pollution problem, because of its spatial extent and the relatively moderate contamination levels. Hence, alternative remediation strategies, such as phytostabilization, are to be investigated. Sustainable phytostabilization projects require the metals to be stabilized in the soil and should thus aim at minimizing metal dispersion via both above- and below-ground pathways. As there is a clear tree species effect on metal uptake, on litter decomposition, on biogeochemical processes in the soil profile and on the water balance, the distribution and fluxes of metals in the forest-soil system will be tree species specific as well. Selecting appropriate tree species is thus crucial for achieving successful phytostabilization. The main objectives of this thesis were to assess the feasibility of phytostabilization by afforesting diffusely Cd and Zn contaminated sandy soils, and to determine the tree species effects on Cd and Zn cycling on these soils. Our study was carried out in ‘Waaltjesbos’, a young post-agricultural forest on a metal contaminated site on sandy soil. At the moment of our study, the trees were 10-15 years old. The six selected tree species were silver birch (Betula pendula), oak (Quercus robur and Q. petraea), black locust (Robinia pseudoacacia), aspen (Populus tremula), Scots pine (Pinus sylvestris) and Douglas fir (Pseudotsuga menziesii). We showed that aspen translocates high amounts of Cd and Zn into its foliage, generating a significant accumulation of total Cd and Zn concentrations in the topsoil (0-5 cm), and this after only 10 years of tree growth. Moreover, accumulated metals in the leaves may pose a long-term risk to primary consumers and enter the food web. Given these risks of above-ground metal dispersion, aspen should be avoided when afforesting Cd and Zn contaminated sites. We also showed that Cd and Zn leaching at 50 cm depth was elevated under black locust, Scots pine and Douglas fir, compared to lower seepage fluxes under silver birch, oak and aspen. This pattern was significantly correlated with leaching of anions and base cations, and to a lesser extent to soil solution pH, at least at this young stage of forest development at a post-agricultural site. Due to former (agricultural) liming and fertilization processes, the base cation status of the soil and the solution pH were still relatively high. However, we expect both parameters to decrease with forest age, especially for species with slowly decomposing litter, implying that the effect of pH on Cd and Zn leaching will become relatively more important in the long term. In this respect, we recommend not to plant tree species that generate high anion, base cation or H+ fluxes when afforesting Cd and Zn contaminated lands, because these species pose a serious risk for metal leaching to deeper soil layers and potential contamination of the groundwater. In general, coniferous as well as N-fixing species should thus be avoided. Finally, contrary to what is often argued in literature about phytostabilization, we found that Cd and Zn leaching losses at our study site showed no resemblance with the downward soil water fluxes. This might imply that the impact of the biogeochemical processes in the soil (solution) on Cd and Zn leaching (complexation with anions, competition for sorption sites with base cations and protons) was relatively more important than the effect of evapotranspiration. Summarized, taking into account all the aforementioned recommendations, it seemed that silver birch and oak may be planted on Cd and Zn contaminated sandy soils. Aspen, black locust, Scots pine and Douglas fir, on the other hand, cause risks for above-ground or below-ground metal dispersion and should therefore be avoided. These conclusions were drawn from our study in a young forest (10 – 15 years). Consequently, it is unsure whether they can be extended to the long term. Further research in older forests is thus essential.