The cyanobacterial hepatotoxin nodularin is abundantly produced by the cyanobacterium Nodularia spumigena in the Baltic Sea during July-August. Nodularin is a potent hepatotoxin and a tumour promoter, distributed in various Baltic Sea environmental compartments, especially food webs involving mussels. Flounders receive nodularin through consumption of blue mussels. In this study nodularin concentrations in individual flounders (liver) were examined between July and September 2002 (six sample sets, four to 10 samples/set), providing information about contribution of sampling on estimates of bioaccumulation intensity. Toxin was determined using liquid chromatography/mass spectrometry (LC/MS) and enzyme-linked immunosorbent assay (ELISA). Additionally, liver histopathology was examined. Observed toxin concentrations were ND-390 μg kg-1 dw (LC/MS) and 20-2230 μg kg-1 dw (ELISA), with maximum concentrations in September (ELISA). The ELISA protocol generally resulted in higher, up to ≈10-fold, toxin concentrations than LC/MS, with increasing difference toward September. This difference may have originated from different extraction solvents in LC/MS and ELISA, ion suppression in LC/MS, and temporal increase in nodularin metabolites detectable with ELISA. The differences in toxin concentrations between individual liver samples were considerable with relative standard deviation values of 20-154% (LC/MS) and 28-106% (ELISA). Since the precision of the ELISA method employed was <25% and that of LC/MS <10%, it can be concluded that the largest source of error in bioaccumulation estimates may be an inadequate number of samples. Although there were tissue lesions in several liver samples, occurrence of lesions was not related to toxin concentrations.