This work tests the previously proposed hypothesis that plant uptake of metals is determined dominantly by diffusional controlled or plant limiting uptake mechanisms at, respectively, low and high metal concentrations. Radish (Raphanus sativus) was grown in 13 soils spiked with Ni (10 and 100 mg kg(-1)) and Cd (0.5 and 4 mg kg(-1)) for 4 weeks to investigate the mechanisms affecting plant uptake. Soil solution concentrations, Css, of Ni and Cd were measured, along with the DGT interfacial concentration, CDGT, and the derived effective concentration in soil solution, CE. Free ion activities, aNi(2+) and aCd(2+), were obtained using WHAM 6. Although there was a poor relationship between Ni in radish roots and either Css or aNi(2+) in unamended soils, the distribution of data could be rationalized in terms of the extent of release of Ni from the soil solid phase, as identified by DGT and soil solution measurements. By contrast Ni in radish was linearly related to CE, demonstrating diffusion limited uptake. For soils amended with high concentrations of Ni, linear relationships were obtained for Ni in radish plotted against, Css, aNi(2+), and CE, consistent with the plant controlling uptake. For Ni the hypothesis concerning dominant diffusional and plant limiting uptake mechanisms was demonstrated. Poor relationships between Cd in radish and Css, aCd(2+), and CE, irrespective of amendment by Cd, showed the importance of factors other than diffusional supply, such as rhizosphere and inhibitory processes, and that fulfilment of this hypothesis is plant and metal specific.