ABSTRACT Crises (e.g. fires, earthquakes, volcanoes) suddenly occur. Some of them may have previous indications, and some may not. The size and impact of a crisis vary depending on its nature, time, place, potential human casualties, capital losses and environmental circumstances. The proposed architecture in this paper provides an approach for an intelligent active DSS that can predict, detect, and manage crises based on integration of a set of efficient technologies. Thus, this architecture employs the technologies of knowledge based systems, real time data base systems, monitoring systems, and special purpose intelligent agents in a way that ensures the reducing, minimizing or eliminating the consequences of crises when they occur. The main components of the proposed architecture, their functions, and the interaction between them are explained. The architecture presented in this paper can be applied as a generic architecture for any intelligent DSS used for managing crises in a wide range of domains and overcome their bad impacts.