Risk assessment is based on scientific information, in part on "regulatory toxicology", i.e., studies following protocols accepted by national or international authorities, and in part on fundamental scientific information clarifying the mechanisms of toxicity and giving a better possibility to evaluate also the findings of routine safety studies. Both are needed, and increased biological understanding increases the possibilities of handling the data in a rational way. In addition, risk assessment seems to include some built-in assumptions that are not necessarily scientific at all. This review attempts to highlight the distortions that are possible if we follow certain rules in a blinkered way not taking into consideration all aspects of the risk. The highly controversial risk assessment of dioxins is used to exemplify the difficulties of excessively straightforward risk analysis process.