Cerium dioxide nanoparticles (CeO2 NPs) are increasingly being used as a catalyst in the automotive industry. Consequently, increasing amounts of CeO2 NPs are expected to enter the environment where their fate in and potential impacts are unknown. In this paper we describe the fate and effects of CeO2 NPs of three different sizes (14, 20, and 29 nm) in aquatic toxicity tests. In each standard test medium (pH 7.4) the CeO2 nanoparticles aggregated (mean aggregate size approximately 400 nm). Four test organisms covering three different trophic levels were investigated, i.e., the unicellular green alga Pseudokirchneriella subcapitata, two crustaceans: Daphnia magna and Thamnocephalus platyurus, and embryos of Danio rerio. No acute toxicity was observed for the two crustaceans and D. rerio embryos, up to test concentrations of 1000, 5000, and 200 mg/L, respectively. In contrast, significant chronic toxicity to P. subcapitata with 10% effect concentrations (EC10s) between 2.6 and 5.4 mg/L was observed. Food shortage resulted in chronic toxicity to D. magna, for wich EC10s of > or = 8.8 and < or = 20.0 mg/L were established. Chronic toxicity was found to increase with decreasing nominal particle diameter and the difference in toxicity could be explained by the difference in surface area. Using the data set, PNEC(aquatic)S > or = 0.052 and < or = 0.108 mg/L were derived. Further experiments were performed to explain the observed toxicity to the most sensitive organism, i.e., P. subcapitata. Toxicity could not be related to a direct effect of dissolved Ce or CeO2 NP uptake or adsorption, nor to an indirect effect of nutrient depletion (by sorption to NPs) or physical light restriction (through shading by the NPs). However, observed clustering of NPs around algal cells may locally cause a direct or indirect effect.