Although phytoextraction using hyperaccumulating plants is seen as a promising technique, a lack of understanding of the basic physiological, biochemical, and molecular mechanisms involved in heavy metal hyperaccumulation prevents the optimization of the phytoextraction technique and its further commercial application. The best-long term strategy for improving phytoextraction is therefore to understand and exploit the biological processes involved in metal acquisition, transport and shoot accumulation in plants. In order to compare Cd allocation in leaves and the role of the leaf cells in hyperaccumulating and non hyperaccumulating plants, we used a wide range of techniques and approaches on contrasting ecotypes of Thlaspi caerulescens, Arabidopsis halleri, Arabidopsis thaliana and one high biomass crop Salix viminalis. As a first step, the extent of hyperaccumulation and tolerance was studied in different plants. Five Swiss T. caerulescens populations were compared for their tolerance and Cd hyperaccumulation to two well studied populations: T. caerulescens Ganges (France) and T. caerulescens Prayon (Belgium). We showed that the behaviour of the populations in hydroponics was linked to the characteristics of their soil of origin. However, growing plants in hydroponics with 1 μM Cd seemed to be adequate to discriminate the various populations tested for hyperaccumulation. Cadmium effect on morphological parameters and Cd accumulation in S. viminalis leaves was also monitored. Salix viminalis was surprisingly tolerant to Cd and high Cd concentration in shoots. Because it may give an indication of the tolerance mechanisms employed by the different plants, we studied the general Cd allocation in leaves of hyperaccumulating (Ganges) and non-hyperaccumulating plants (Prayon, S. viminalis). Surprinsingly, when grown in hydroponics the only differences found between Prayon and Ganges at the leaf level was concentration of Cd found in the storage sinks. Results also showed similarities between T. caerulescens and S. viminalis in Cd storage, although the Cd concentrations found in leaves differed greatly. Both point-like accumulation and accumulation at the edges of the leaves were observed. A link could be established with visible symptoms (necrosis). At last differences in Cd localization between young and mature leaves were observed. These differences were attributed to physiological differences between leaves. Salix viminalis and T. caerulescens were both able to reduce Cd toxicity by allocating Cd in less sensitive tissues. Cadmium was found inside the cells and in the cell walls of the leaves of T. caerulescens, but mainly in cell walls and to a lesser extent in the symplasm in S. viminalis. Metal allocation in both plants indicated that the plant general growth strategy governed metal accumulation. We further investigated the role of the leaf cells in allocating Cd in leaves of Ganges, A. halleri and Prayon by characterizing Cd uptake in mesophyll protoplasts. Results indicated that differences in metal uptake could not be explained by different constitutive transport capacities at the leaf protoplast level. However, pre-exposure of the plants to Cd induced an increase in Cd accumulation in protoplasts of Ganges, whereas it decreased Cd accumulation in A. halleri protoplasts. The experiment with competitors eventually showed that probably more than one single transport system are carrying Cd in parallel into the cell. Metal allocation indicates that the principle of metal storage in metabolically less sensitive plant parts governs metal accumulation. Vacuolar compartmentalization and cell wall binding in leaves could therefore both play a role in accumulation of heavy metals. Based on these various results, we suggested that metal storage in plant demands the involvement of more than one compartment. Further work is however needed to assess many steps of the trafficking of metals that remain enigmatic.