Presenting the results of an ambitious research activity this book intends to understand why Early Warning Systems (EWSs) fail. However, from the beginning, the objective turned out to be challenging; first, because, so far, there is not a shared understanding of what an EWS is (among both researchers and practitioners communities); second, as a consequence, because it is equally unclear when an EWS can be considered successful or not. Due to this, the research needed first to face some “open questions” instead of going straight to the point under investigation. Specifically, it was first necessary to define EWS, identify its components and functions, peculiarities, and weak points. Only at that point, a first attempt to evaluate EWS performance was possible. Flood Early Warning Systems Performance is organised according to the conceptual steps required by the research. In part I the “open questions” about the definition and the role of EWSs are handled, the aim being the identification of how to evaluate their effectiveness/performance. Part II focuses on the real aim of the research, providing concepts and tools to assess EWS performance; suggested tools are also implemented in a case study to describe how they can be applied in practice. Focusing specifically on the topic of flood risk in mountainous regions, the book can be viewed as a sort of manual for EWS designers, managers, and users. It is organised into different independent sections which will appeal both to experts as well as those with an interest in the subject. Most of results can also be exported to other hazards.