One of the objectives of the early warning and surveillance systems established by the French Institute for Public Health Surveillance is to detect infectious threats in order to mitigate their impact. For potentially emerging infectious diseases, detection and risk analysis present unique challenges due to the inherent uncertainty. Numerous signals are captured by surveillance systems. These concern new agents, agents that are little known because they are not endemic to the region, or abnormal increases in cases of a known disease. These signals are verified, analyzed, and then classified as a potential threat to public health, a signal requiring monitoring, or no threat. Using various examples, we illustrate the method and criteria for analyzing risk and classifying reported phenomena as emerging, emphasizing the importance of the concept of exposure among certain at-risk population groups, such as livestock workers, healthcare providers, travelers, immunocompromised individuals, etc. This method should improve responsiveness, the quality of expert analysis, and feedback regarding these initially vague phenomena, and better assess future surveillance needs. (R.A.)