Fish from Dutch markets were analysed for concentrations of polychlorinated dibenzo-p-dioxins/dibenzofurans (PCDD/Fs) and dioxin-like polychlorinated biphenyls (DL-PCBs) and compared with the new European maximum residue levels (MRLs), set in 2006. In a first study on 11 different fish and shellfish from various locations, concentrations of PCDD/Fs were nearly all below the MRL for PCDD/Fs [4 pg toxic equivalents (TEQ) per gram wet weight (ww)] and nearly all below 8 pg total TEQ/g ww, the new MRL for the sum of PCDD/Fs and DL-PCBs. Some samples exceeded the total TEQ MRL, such as anchovy, tuna and sea bass. Furthermore, 20 (out of 39) wild eel samples exceeded the specific MRL for eel (12 pg total TEQ/g ww), as the study revealed PCDD/F TEQ levels of 0.2-7.9 pg TEQ/g ww and total TEQ values of 0.9 to 52 pg/g ww. TEQ levels in farmed and imported eel were lower and complied with the MRLs. Smoking eel, a popular tradition in the Netherlands, only had marginal effects on PCDD/F and DL-PCB concentrations. Owing to volatilization, concentrations of lower-chlorinated PCBs were reduced to below the limit of quantification after smoking. DL-PCBs contributed 61-97% to the total TEQ in all eel samples. This also holds for other fish and shellfish (except shrimps): DL-PCB contributed (on average) from 53 (herring) to 83% (tuna) to the total TEQ. Principal-component analysis revealed distinctive congener profiles for PCDD/Fs and non-ortho PCBs for mussels, pikeperch, herring and various Mediterranean fish. The application of new TCDD toxic equivalency factors (TEFs) set by the World Health Organization in 2006 (to replace the 1997 TEFs) resulted in lower TEQ values, mainly owing to a decreased mono-ortho PCB contribution. This decrease is most pronounced for eel, owing to the relative high mono-ortho PCB concentrations in eel. Consequently, a larger number of samples would comply with the MRLs when the new TEFs are applied. The DR CALUX(R) assay may be used for screening total TEQ levels in eel, in combination with gas chromatography-high resolution mass spectrometry confirmation of suspected samples. An almost 1:1 correlation was found when the 1997 TEFs were applied, but, surprisingly, a 1.4-fold overestimation occurred with application of the 2006 TEFs.