During 2002-2003 a dredging campaign was undertaken in Göteborg harbor, Sweden, to widen and deepen shipping lanes. A bunker oil spill occurred in the harbor in July 2003, thereby further exasperating the situation for marine life. Eelpout, Zoarces viviparus, was used as a sentinel species to monitor the impact of these events. Here, we have investigated the effects on two liver parameters, lipid peroxidation and protein carbonylation, which can serve as biomarkers for oxidative stress. Lipid peroxidation data, measured as TBARS, in eelpout from the field study showed no significant differences between reference and polluted sites, at any of the time points. These results are mirrored in a laboratory exposure to the bunker oil where no differences were seen between the control and the exposure groups. A trend towards a seasonal cycle in TBARS levels in eelpout liver was observed, with highest amounts measured during the warmer months, 97.08+/-14.45 nmol/g liver in the harbor in July compared to 41.20+/-2.66 nmol/g liver in November 2001. Protein carbonylation, measured using an ELISA method, did, however, show differences between the reference and polluted sites in the field, as well as differences between time periods (before and during dredging and following the oil spill). The laboratory exposure indicated that the formation and/or accumulation of protein carbonyls is greatly affected by exposure to this PAH rich oil. Levels in the control group were 1.76+/-0.13 nmol/mg protein while those fish exposed to the high dose had 6.23+/-0.17 nmol/mg protein. We concluded that TBARS is not an appropriate biomarker for pollutant mediated oxidative damage in eelpout while protein carbonyl formation does appear to be affected by xenobiotic exposure.