Planktothrix rubescens, the dominant cyanobacterium in Lake Zürich, is generally considered to be toxic to zooplankton. The major toxin was determined by NMR spectroscopy and chemical analysis to be [D-Asp3,(E)-Dhb7]microcystin-RR. The compound was isolated in high purity, and its 24-h acute grazer toxicity was compared with microcystin-LR, microcystin-RR, microcystin-YR, and nodularin using a Thamnocephalus platyurus bioassay. Based on LC50 values D-Asp3,(E)-Dhb7]microcystin-RR was the most toxic microcystin tested. Nodularin was slightly more toxic under the conditions of the assay. The large number of individuals available for the grazer bioassay allowed the determination of dose-response curves of the different microcystins. These curves showed marked differences in their steepness. Microcystin-RR, which had nearly the same LC50 as microcystin-LR and microcystin-YR, exhibited a very flat dose-response curve. This flat curve indicates that, for some individuals, lower concentrations of this microcystin are much more toxic than are the other two microcystins. Mortality of 100% requires much higher concentrations of microcystin-RR, indicating the resistance of some animals to the toxin. The purified D-Asp3,(E)-Dhb7]microcystin-RR exhibited a higher molar absorption coefficient determined by quantitative amino acid analysis than the coefficients generally used for other microcystins. This observation has consequences for the risk assessment for microcystins and makes a structural determination of microcystins an absolute requirement. The presence of the dehydrobutyrine residue may be the reason for the higher specific toxicity of D-Asp3,(E)-Dhb7]microcystin-RR when compared to the N-methyldehydroalanine-containing microcystins.