Available risk analysis techniques are well adapted to industry since they were developed for its purpose. All hazards met in industry are also present in research/academia (although quantities of some hazardous substances are smaller). Still, because of its characteristics (high turnover of collaborators, rapid reorientation of research programs, freedom of research, equipment often in development stage, difficulty to obtain accidents statistics, not well-established processes, etc.), research/academia is an environment whose risks are difficult to analyze by conventional techniques. In this paper, we discuss various risk analysis methods in the light of their adequacy for an academic environment. Finally, we propose the bases for the development of a new risk analysis methodology for complex areas such as academia/research. This method should be fast, intuitive, semi quantitative, and easy to use. It should lead to risk-ranking providing the identification of critical areas and prioritization of safety actions. It is suggested to estimate this ranking as a combination of severity, probability of accident, hazard detectability, and worsening factors (where academia/research specificities are considered). The formula, in which risk is a nonlinear function of the constituting elements, will have a general form allowing getting a risk index for all hazard categories in the same manner.