Zebra mussels (Dreissena polymorpha) are an important component of the foodweb of shallow lakes in the Netherlands, amongst others in Lake IJsselmeer, an international important wetland. Large numbers of ducks feed on these mussels in autumn and winter. The mussels are filter feeders and are exposed to high densities of cyanobacteria in summer and autumn. Mussels and cyanobacteria both thrive in Lake IJsselmeer. Apparently the mussels are somehow protected against accumulation of harmful quantities of cyanobacterial toxins. In this study, we investigated the assimilation of the cyanobacterial toxin microcystin-LR (MC-LR) in zebra mussels when fed the toxic cyanobacterium Microcystis aeruginosa as sole food or in a mixture with the eustigmatophyte Nannochloropsis limnetica. After 3 weeks of assimilation we studied the depuration of MC-LR during 3 weeks when the food of the mussels was free of cyanobacteria. These assimilation/depuration experiments were combined with grazing experiments, using the same food treatments. Microcystins were analyzed using liquid chromatography-mass spectrometry (LC-MS); in addition, covalently bound MC were analyzed using the MMPB method. The mussels showed higher clearance rates on Microcystis than on Nannochloropsis. No selective rejection of either phytoplankton species was observed in the excretion products of the mussels. Zebra mussels fed Microcystis as single food, assimilated microcystin-LR relatively fast, and after 1 week the maximum value of free unbound microcystin assimilation (ca. 11 μg g DW -1) was attained. For mussels, fed with the mixed food, a maximum of only 3.9 μg g DW -1 was recorded after 3 weeks. Covalently bound MC never reached high values, with a maximum of 62% of free MC in the 2nd week of the experiment. In the depuration period microcystin decreased rapidly to low values and after 3 weeks only very low amounts of microcystin were detectable. The amount of toxin that accumulated in the mussels would appear to be high enough to cause (liver) damage in diving ducks. However, death by exposure to microcystin seems unlikely. Mussels seem efficient in minimizing the assimilation of microcystin. If it were not for this, mass mortalities of ducks in shallow lakes in the Netherlands would presumably occur on a much more widespread scale than is currently observed.