Scientific and technological advances which took place during the twentieth century have totally changed the relationship between science and society and, as a consequence, the relationship between science and law, that are nowadays more and more involved in several fields of the governmental action. Public authorities working at a local, a supranational and a global level, are often in charge of regulating the risks which arise also from new technologies; therefore, today, many policy and decision-making processes enshrine a scientific or technological dimension that often lead policymakers to seek scientific advice, in order to provide public policies and decisions related to risks with a solid foundation and legitimation. Science-based decision-making processes are particularly relevant in the European Union context, where “the best available science” becomes a key input in many decisions adopted by EU institutions, in particular in the food safety domain, that is somewhat paradigmatic with regard to the issue of risk regulation in the EU. The outbreak of the 1996 BSE crisis, and other food scandals, have shown the inadequacy of the former EU approach to food safety regulation, as applied until then, and called for a reform of the system, that lead to the enactment of regulation (EC) n. 178/2002 and the establishment of the European Food Safety Authority (EFSA). The cornerstone of the new global and science-based EU food policy is the risk analysis scheme, structured upon three different components: risk assessment, risk management and risk communication. In the dichotomy between risk assessment and risk management, respectively entrusted to EFSA and to the European Commission, we can find one of the key features of this legislation, with the aim to ensure the excellence and, first and foremost, the independence of the scientific outputs from political influences, on the one hand, and providing political authorities with sound scientific basis for their regulatory choices, on the other. In practice, a clear-cut distinction between risk assessment and risk management is nevertheless problematic, and this has been demonstrated by the concrete way in which one of the most important, and also contested, regulatory fields (like GMOs’ regulation) has worked until today. In the context of the GM food and feed authorization procedure established at the EU level there is a close interaction between risk assessment and risk management spheres, resulting in a relevant influence by EFSA on the Commission risk management decisions. In several cases, the European Commission showed a great reliance on scientific information and advice deriving from risk assessment conducted by EFSA. The role of EFSA in this framework is however strongly contested. EFSA is often blamed of not being really and totally independent from political and economic interests, with consequential problems with regard to the recognition of the Authority as the legitimacy provider of the measures adopted by the European Commission in the health and in the food safety domain. Risk management measures mostly consist in a balance between opposite interests and values; this balance involved a certain discretion by the political authorities, which is particularly wide in cases where scientific uncertainty requires to behave on a precautionary basis, so granting a sort of “advanced” protection. However, the exercise of this discretion have to be balanced with the results of the scientific risk assessment, able to granting the (technical and scientific) legitimation needed for the risk management activity. The complex balance between the technical and scientific moment, and the political one, is the heart of our matter, namely of the risk regulation process; behind it, there is a tension between the functional necessity for science-based decision-making and the wider demand for some kind of public participation; this tension is epitomized in the dichotomy “input-output legitimacy”. The EU risk regulation in the food domain is therefore paradigmatic in studying and in understanding the wide and complex questions related to the legitimation of the EU decisionmaking processes (and of the European Union itself), still often perceived (although less and less) as being affected by a legitimacy and democratic deficit.