A summary of a critical review by a working group of the German Federal Environment Agency and the German Federal Institute for Risk Assessment on the carcinogenic potential of nanomaterials is presented. After a critical review of the available data, we conclude that the potential carcinogenic risk of nanomaterials can currently be assessed only on a case-by-case basis. There is certain evidence that different forms of CNTs (carbon nanotubes) and nanoscale TiO(2) particles may induce tumours in sensitive animal models. It is assumed that the mode of action of the inhalation toxicity of asbestos-like fibres and of inhalable fractions of biopersistent fine dusts of low toxicity (nano-TiO(2)) is linked to chronic inflammatory processes. Existing epidemiological studies on carcinogenicity for these manufactured nanomaterials are not sufficiently conclusive. Generally speaking, the database is not adequate for an assessment of the carcinogenic potential of nanomaterials. Whereas a number of studies provide evidence of a nano-specific potential to induce tumours, other studies did not. This is possibly due to insufficient characterisation of the test material, difference in the experimental design, the use of different animal models and species and/or differences in dosimetry (both with regard to the appropriate dose metric and the estimated effective dose quantities). An assessment of the carcinogenic potential and its relevance for humans are currently fraught with uncertainty. Furthermore, the nano-specificity of the carcinogenic effects observed cannot be conclusively evaluated. Specific carcinogenic effects of nanomaterials may be both quantitative and qualitative. In quantitative terms, the carcinogenic effects of nanoparticles are thought to be simply more pronounced compared to the corresponding bulk material (due, for example, to the considerably larger surface area and higher number of particles relative to the mass concentration). On the other hand, certain nano-properties such as small size, shape and reactivity, retention time and distribution in the body after overcoming biological barriers, as well as subcellular and molecular interactions may play a role in determining the toxicity in qualitative terms, i.e. the carcinogenic potential of the nanomaterial and the non-nanoscale comparison substance may be fundamentally different. All of these factors leave no doubt about the fact that there is a great need for research in this area and that new standardised test methods need to be developed or existing ones adapted at the very least to achieve valid answers regarding the carcinogenic potential of nanomaterials. Global production of nanomaterials is set to increase in the years to come, and new materials with new properties will be developed, so that greater human exposure to them must be anticipated. No reliable conclusions can currently be drawn about exposure to nanoparticles and their release from products. Firstly, there are substantial deficits in information about the processing of nanomaterials in products and preparations. Secondly, there are only a small number of studies on nanoparticle release, and reliable techniques for measuring and monitoring nanomaterials in different environmental media are still being developed which is both complex and costly. Despite the uncertainties, the findings to date on the carcinogenic potential of nanomaterials must be taken seriously, and precautionary measures to minimise exposure should go hand in hand with the development of a comprehensive and conclusive toxicological methodology and testing procedure for nanostructured materials that includes all possible exposure routes. With regard to possible legal classification of nanomaterials and the transferability of classifications of their non-nanomaterial counterparts, we believe it is necessary to have separate procedures for nano and non-nano forms. Furthermore, criteria for evaluating nano-specific carcinogenic properties should be constantly updated and adapted to the state of knowledge. There is a need here for amendments to be made to EU legislation, as currently nanoforms do not represent a separate category of substance in their own right.