The impact of the environment (air, water, food pollution) on health is a major concern in contemporary society. Unfortunately, there are relatively few objective epidemiological data on this subject and their accuracy is limited. Risks are often not quantified, whereas in public health the quantitative assessment of the various risks and benefits must provide the bases for a global strategy. Actual risks should be distinguished from putative risks and, when the risks are putative, an effort should be made to ascertain the upper and lower limits of the risk. The validity of a linear no threshold relationship for assessing putative risks should be discussed and, whenever appropriate, other relationships should be considered. Since emotional reactions often pervade environmental issues, which in turn are exploited for political or commercial reasons, it is not surprising that any statement or action may provoke violent debate. It is serious to underestimate the importance of a risk, since appropriate measures may not be put in place. However, it is equally serious to overestimate it because this can provoke unjustified fears, a pervasive unease, and a rejection of certain technologies, even to the point of discrediting science. It can lead therefore to a questioning of progress by instilling fears about any innovation, as well as facilitating the manipulation of public opinion for financial or ideological reasons, and finally to distortions in budget allocations and public health actions. Confronted with this situation, the Academy’s role should be threefold. a) Whenever necessary, point out the need for an increase in appropriate fundamental research. When epidemiological data are uncertain, analyse the cause of these uncertainties and advocate appropriate development in statistical methodologies and epidemiological research, which could ascertain the upper limit of the putative risk. The lack of knowledge often results in public anxiety; this reaction should be investigated and psychosociological research must be encouraged and supported. b) Inform the scientific community and the public; fight against misinformation and sensationalism in the news, and take advantage of the Internet to this end. Encourage openness and transparency in the preparation of scientific reports and dialogue with the stakeholders. c) Better define the role and the place of experts, ensure their independence, monitor their competence and make sure they represent the various fields involved. When reports are conflicting, the Academy should be ready to organize a forum for analysing the roots of these disagreements and to delineate the limits of the uncertainties.RésuméL’impact de l’environnement sur la santé constitue une préoccupation majeure dans le monde contemporain. Malheureusement les données objectives, épidémiologiques sur ce sujet sont relativement peu nombreuses et leur précision est limitée. Étant donné les réactions émotives qui entourent ces problèmes et l’exploitation qui en est souvent faite pour des raisons commerciales ou politiques, il n’est pas surprenant que tout résultat et toute action suscitent des polémiques parfois violentes. Or il est aussi grave de sous-estimer l’importance d’un risque, puisque l’on ne met pas en œuvre les mesures nécessaires, que de le surestimer, ce qui peut d’une part provoquer des craintes injustifiées, une atmosphère d’inquiétude, un rejet de certaines technologies, voire un discrédit de la science, une remise en question du progrès et la crainte de toute innovation, et d’autre part faciliter les manipulations de l’opinion pour des raisons financières ou idéologiques et entraîner des distorsions dans les allocations budgétaires et les actions de santé publique. Devant cette situation, le rôle des Académies devrait être triple : a) Accroître les connaissances fondamentales, développer des méthodologies statistiques et épidémiologiques, soutenir les recherches psychosociologiques sur les risques. b) Informer la communauté scientifique et le public, lutter contre la désinformation et la sensationalisation des nouvelles. c) Mieux définir le rôle et la place des experts, assurer leur indépendance, veiller à leur compétence et à leur représentativité.