Financial crises have large deleterious effects on economic activity, and as such have been the focus of a large body of research. This study surveys the existing literature on financial crises, exploring how crises are measured, whether they are predictable, and why they are associated with economic contractions. Historical narrative techniques continue to form the backbone for measuring crises, but there have been exciting developments in using quantitative data as well. Crises are predictable with growth in credit and elevated asset prices playing an especially important role; recent research points convincingly to the importance of behavioral biases in explaining such predictability. The negative consequences of a crisis are due to both the crisis itself but also to the imbalances that precede a crisis. Crises do not occur randomly, and, as a result, an understanding of financial crises requires an investigation into the booms that precede them.