Flood damages in the United States continue to increase despite increased spending for flood control and mitigation, suggesting that new methods for estimating and characterizing flood hazards would be valued by society. Paleoflood hydrology was developed in the 1970s as a way to understand the magnitude of extreme flooding in central Texas and has evolved considerably in the last 30 years to become an important scientific endeavor with broad scientific and social relevance. The interdisciplinary methods of paleoflood hydrology have evolved to overcome difficulties, real and perceived, involving (1) inaccuracies in estimating the ages of floods, (2) inaccuracies in reconstructing flood discharges, (3) lack of robust statistical methods for incorporating paleoflood data in flood-frequency analysis, and (4) the effects of climatic shifts and nonstationarity. Of these, only the fourth represents a real problem that is faced by both conventional flood hydrology and paleoflood hydrology. Its resolution requires further collection of flood data, both from gaging stations on unregulated rivers and paleoflood sites. The basic techniques of paleoflood hydrology lead to the use of paleoflood data in flood-hazard analysis, both in terms of traditional flood-frequency analysis and the potential use of flood history within the context of flood-hazard assessment. Because paleoflood data reduce the uncertainty in estimates of long return-period floods, use of this technique offers substantial societal benefit, particularly in the design or retrofitting of dams or other significant floodplain structures. Improving the general understanding of hydroclimatic effects on flood frequency represents one of the most important contributions that paleoflood hydrology can make in the future. With continued methodological and conceptual refinement, paleoflood hydrology has the potential to guide important changes in the traditional paradigms of flood-frequency analysis and flood-hazard assessment.