a b s t r a c tBackground: Microcystins (MCs) contaminate water bodies due to cyanobacterial blooms all over the world, leading to frequent exposure of humans to MCs through consumption of meat, fish, seafood, blue green algal products and water, accidental ingestion of contaminated water and scum during recre ational activities and inhalation of cyanobacterial aerosols. For monitoring of human exposure, sensitive screening methods are needed. However, during the analytical process of various matrices, such as human serum, some problems appear to regularly occur during sample preparation and storage, leading to MC loss and thus to underestimation of the true MC concentration. The aim of the current study was therefore to assess the pitfalls of the MC extraction method from human serum with more detail. Methods: Six MC congeners (MC LR, YR, RR, LA, LW, LF) and defined equimolar MC mixtures thereof were spiked into human serum, and quantified using the commercially available Adda ELISA subsequent to standard extraction (methanol extraction with subsequent SPE). To detect the potential influence of sample storage and preparation/storage materials, different types of material such as glass, standard polypropylene and surface treated polypropylene were compared. Results: Loss of MC during preparation and storage is largely dependent on (1) the handling of the stored material, (2) the 'surface' of the storage material and (3) the hydrophobicity of the MCs. Conclusions: The pitfalls described for MC analysis with the ELISA are primarily associated with sample preparation and clean up and thus also apply to other analytical techniques for MC detection beyond the ELISA used. It can be concluded that ELISA based methods are suitable tools for the detection of MCs in human sera and other samples.