This work set out to test the hypothesis that uptake of metals by hyperaccumulator (HA) plants is more likely to be diffusion limited than uptake by nonhyperaccumulator (NHA) plants. Two circumneutral soils, with different contents of organic matter (0.8 and 5.8%), were amended with Cd (0.5 to 5 mg kg(-1)) and Ni (10 to 100 mg kg(-1)). A Cd HA plant, Thlaspi caerulescens, was grown in pots containing the Cd amended soils, and a Ni HA, Thlaspi goesingense, was grown in pots containing the Ni amended soils. A NHA plant of the same family, Thlaspi arvense, was grown in the same soils. Metals were measured in both roots and shoots of all plants. Concentrations of Cd and Ni were measured in soil solution and using the technique of diffusive gradients in thin-films (DGT). The dependencies of metal measured by DGT, [M]DGT, and in soil solution, [M]ss, on the amended metal concentration, [M]add, were consistent with fast supply of Cd but a slower rate of release of Ni from solid phase to solution at lower [Ni]add. Detailed consideration of the dependence of Ni and Cd in shoots and roots on [M]add, [M]ss, and [M]DGT allowed assessment of the supply mechanism. The weight of evidence suggested that diffusion limitation applies for uptake of Cd by both HA and NHA plants and for uptake of Ni by the HA. However, uptake of Ni by the NHA is not limited by diffusion and the biotic ligand model is probably appropriate.