Kinetic parameters (uptake from solution and elimination rate constants) of Cu, Ni and Pb bioaccumulation were determined from two Gammarus pulex and three Gammarus fossrum wild populations collected from reference sites throughout France in order to assess the inter-species and the natural inter-population variability of metal bioaccumulation kinetics in that sentinel organism. For that, each population was independently exposed for seven days to either 2.5 mu g L-1 Cu (39.3 nM), 40 mu g L-1 Ni (681 nM) or 10 mu g L-1 Pb (48.3 nM) in laboratory controlled conditions, and then placed in unexposed microcosms for a 7-day depuration period. In the same way, the possible influence of metal exposure history on subsequent metal bioaccumulation kinetics was addressed by collecting wild gammarids from three populations inhabiting stations contaminated either by Cd, Pb or both Pb and Ni (named pre-exposed thereafter). In these pre-exposed organisms, assessment of any changes in metal bioaccumulation kinetics was achieved by comparison with the natural variability of kinetic parameters defined from reference populations. Results showed that in all studied populations (reference and pre-exposed) no significant Cu bioaccumulation was observed at the exposure concentration of 2.5 mu g L-1, Concerning the reference populations, no significant differences in Ni and Pb bioaccumulation kinetics between the two species (G. pulex and G. fossarum) was observed allowing us to consider all the five reference populations to determine the inter-population natural variability, which was found to be relatively low (kinetic parameters determined for each population remained within a factor of 2 of the minimum and maximum values). Organisms from the population exhibiting a Pb exposure history presented reduced Ni uptake and elimination rate constants, whereas no influence on Ni kinetic parameters was observed in organisms from the population exhibiting an exposure history to both Ni and Pb. Furthermore Pb bioaccumulation kinetics were unaffected whatever the condition of pre-exposure in natural environment. Finally, these results highlight the complexity of confounding factors, such as metal exposure history, that influence metal bioaccumulation processes and showed that pre-exposure to one metal can cause changes in the bioaccumulation kinetics of other metals. These results also address the question of the underlying mechanisms developed by organisms to cope with metal contamination.