This study aimed to increase the knowledge about the toxicity of fish-derived organohalogen pollutants in mammals. The strategy chosen was to separate organohalogen pollutants derived from Baltic herring (Clupea harengus) fillet, in order to obtain fractions with differing proportions of identified and unidentified halogenated pollutants, and to perform a subchronic toxicity study in rats, essentially according to the OECD guidelines, at three dose levels. Nordic Sea lodda (Mallotus villosus) oil, with low levels of persistent organohalogen pollutants, was used as an additional control diet. The toxicological examination showed that exposure to Baltic herring oil and its fractions at dose levels corresponding to a human intake in the range of 1.6 to 34.4 kg Baltic herring per week resulted in minimal effects. The spectrum of effects was similar to that, which is observed after low-level exposure to pollutants such as chlorinated dibenzo-p-dioxins and dibenzofurans (CDD/F) and chlorinated biphenyls, despite the fact that these contaminants contribute to a minor part of the extractable organically bound chlorine (EOCI). The study confirmed previous findings that induction of hepatic ethoxyresorufin deethylase (EROD) activity takes place at daily intake levels 0.15 ng fish-derived CDD/F-TEQs/kg body weight. The study also demonstrated that hepatic vitamin A reduction takes place at somewhat higher daily exposure levels, i.e. 0.16-0.30 ng fish-derived CDD/F-TEQs/kg body weight. Halogenated fatty acids, the major component of EOCI, could not be linked to any of the measured effects. From a risk management point of view, the study provides important new information of effect levels for Ah-receptor mediated responses following low level exposure to organohalogen compounds from a matrix relevant for human exposure.