This research aimed to estimate potential genotoxicity for consumers resulting from the ingestion of seafood contaminated with polycyclic aromatic hydrocarbons (PAHs) released into the marine environment after the 'Erika' shipwreck along the coasts of south Brittany, in France. Mussels (Mytilus sp.) collected from sites on the Atlantic coast that were affected by the oil slick in various degrees, were used to feed rats daily for 2 and 4 weeks. DNA damage was measured by use of the Comet assay in the liver, bone marrow and blood of rats receiving food contaminated with 312 microg of 16PAHs/kg dry weight (d.w.) equivalent to 33.8 microg TEQs (toxic equivalent quantities to benzo(a)pyrene (BaP))/kg d.w. mussels, 569 microg/kg d.w. (83.6 microg TEQs/kg) and 870 microg/kg d.w. (180.7 microg TEQs/kg). A dose-effect-time relationship was observed between the amount of DNA damage in the liver and bone marrow of the rats and the PAH contamination level of the mussels. Genotoxicity increased during the period between 15 and 30 days in rats that received food at the highest two PAH levels. On the other hand, no significant change in liver and bone marrow of rats fed with mussels containing 33.8 microg TEQs/kg d.w. was recorded at 30 days compared with 15 days, indicating efficient DNA repair capacity at low levels of exposure. No signs of genotoxicity were found in peripheral blood. Globally, the observed effects were rather moderate. These results show that oil-contaminated food caused DNA damage in predators, and underline the bioavailability to consumers of pollutants in mussels contaminated with fuel oil. The usefulness of the Comet assay as a sensitive tool in biomonitoring studies analyzing responses of PAH transfer through food webs was also confirmed.