A recolonisation experiment was performed using sediments from three locations (Nords Wharf, Cockle Bay and Warners Bay) along a metal contamination gradient (Lake Macquarie, Australia). The study aimed to determine whether the source of the sediments would influence the recolonisation of benthic assemblages, providing additional information regarding the ecological risks associated with the more contaminated sediments. Sediments were translocated to two locations within the lake and retrieved after 22 weeks along with benthic samples from the surrounding sediments (ambient). Total abundance was greater in the reference treatment (Nords Wharf), with this difference being driven by polychaetes, especially capitellids. In general, univariate metrics were similar among the recolonised treatments, although evenness and diversity patterns were complex due to significant location-treatment interactions. PERMANOVA analysis demonstrated that the Nords Wharf treatments were significantly different from the more contaminated treatments (Cockle Bay and Warnerâ¿¿s Bay) and the ambient assemblages, with no differences being detected among Cockle Bay and Warners Bay assemblages. Collectively, the findings showed that the source of the sediments influenced the composition of the recolonised assemblages, with the described approach being a powerful tool for examining the effects of location-specific sediments under environmentally relevant conditions.