The accident that occurred in Seveso in 1976 changed our approach to risk assessment forever. For instance, it led to drafting specific European directives to prevent major accident hazards. Their focus is on industrial sites handling dangerous substances and denominated "Seveso sites". Their operators have the obligation to provide specific information to the competent authorities, such as safety reports. However, risk analysis studies usually provide a static picture of the site status, while the system constantly evolves or degrades. For this reason, this contribution suggests a dynamic risk analysis approach aiming to continuously calibrate and improve, based on new related evidence and lessons learned. The work suggests focusing on such early deviations to lower the probability of high impact low probability (HILP) events. Three complementary methods may be used to process such information: dynamic hazard identification, dynamic analysis of initiating events, and dynamic analysis of consequences. A representative example of their potential is provided by comparing their capabilities with the causes that led to the Seveso catastrophe. Despite its limitations, Dynamic Risk Analysis represents an opportunity for improved decision-making support and critical risk communication. (C) 2017 Elsevier Ltd. All rights reserved.