Sublethal exposures of the marine amphipod Gammarus locusta to a concentration range of copper (Cu) in water (4 days' exposure; 3, 5 and 10 micro g Cu l(-1)) or spiked sediments (28 days' exposure; 1, 3 and 6 mg Cu kg(-1) dry weight) were performed, and the resulting bioaccumulation of Cu and effects on putative metallothionein (MT) and lipid peroxidation (LP) were investigated. A time-course exposure study (over 10 days) to a single water-borne concentration of Cu (4 micro g l(-1)) was also carried out. MT and LP were quantified, respectively, by differential pulse polarography and as thiobarbituric acid-reactive malondialdehyde equivalents. The increasing levels of Cu in water and sediment exposures resulted in enhanced uptake of the metal by G. locusta. Synthesis of putative MT occurred in response to exposure to water-borne Cu, the levels being higher (p < 0.05) over the dose range of Cu compared with controls. A positive correlation was observed between putative MT levels and the Cu body-burden concentration (p < 0.001). However, no increase in LP was observed in these animals. In contrast, in the time-course experiment, LP levels increased within 1 day of exposure, subsequently peaking at 4 days (68% greater than control, p < 0.001), before returning to control values by day 6. Higher levels of MT were also observed in this exposure, but at days 6 and 10 (55% and 38%, respectively), paralleling the decrease in LP. No increase in MT levels was recorded with exposure to Cu-contaminated sediments, whereas higher levels of LP were seen in comparison with controls (p < 0.001). Overall, the inverse relationship between putative MT induction and the occurrence of LP indicates that MT may protect against the prooxidant effects of Cu. It is concluded that MT and LP offer potential for application as biomarkers in G. locusta.