The technique of diffusive gradients in thin films (DGT) binds metal ions to a resin after they have diffused through a well-defined layer of gel. If metal complexes dissociate in this diffusion layer, they will in principle be measured. By varying the thickness of the layer the extent of metal dissociation can be controlled. These principles were used to examine the lability of metal complexes, and, where possible, to determine the dissociation rate constant. DGT measurements were made using devices with a range of diffusive layer thicknesses (0.16-2.0 mm) in solutions containing copper or nickel in the absence and presence of nitrilotriacetic acid (NTA). Rate expressions were derived to relate the transfer kinetics to the effective measurement time of DGT. The dependence of DGT measurements on diffusion layer thickness can be modeled assuming a dissociation rate constant, k-1, of 3.6 ± 0.5 × 10-5s-1 for NiNTA-. CuNTA- was found to be fully labile, with k-1 > 0.012 s-1. The results agreed well with the limited rate data available in the literature. This work has demonstrated, for the first time, the validity of the assumption that only the free metal ion and not the metal complex, reacts with the binding resin of the DGT device. DGT therefore has the potential to distinguish between adjunctive and disjunctive mechanisms of complex dissociation. Because DGT can be readily deployed in situ, in natural waters, soils, and sediments, it opens up the possibility of directly obtaining kinetic information in natural or contaminated environmental systems.