Mussels (Mytilus edulis) transferred in net bags from clean to chronically mercury polluted water readily accumulated mercury during an exposure period of three months. Growth of the transplanted mussels had a “diluting” effect on the mercury concentration, but the absolute weight of mercury uptake increased throughout the entire period, though there was a tendency for decreased efficiency of the removal of mercury per liter of water filtered by the mussels. Mussels were also translocated from polluted to clean (laboratory) water to depurate mercury. The biological half-lives of mercury was 293 d for M. edulis from the chronically polluted area in contrast to only 53 d for mussels from a temporary massive mercury polluted area near a chemical deposit. In both cases about 75% of the total mercury in the mussels was inorganic, and it is suggested that both inorganic and organic mercury species were immobilized in mussels from the long-term mercury polluted area, whereas the immobilization capacity was exceeded in the short-term mercury exposed mussels near the chemical deposit. Very slow elimination of mercury was observed in the deposit-feeding bivalve Macoma balthica from the chronically polluted area, and about 6% of the total mercury was methyl-+phenyl-mercury. This is more than three times lower than found in M. edulis from the same collecting site. A pronounced difference in the mercury speciation (i.e., total mercury, total organic mercury, methyl-mercury and phenyl-mercury) in M. edulis from the two mercury polluted areas is thought to reflect the different character of the mercury pollution in the two areas.