Losses resulting from natural hazards floods, hurricanes, earthquakes, tornadoes, wildfires cost the U.S. economy billions of dollars each year. Moreover, the costs are escalating, in large part because of the growing population in coastal and other high-risk areas. Further, the increasing complexity of the nation's infrastructure, particularly in urban areas, means that potential losses during a natural hazard are only likely to increase. As part of its strategy to address the hazard loss problem, the federal government funds research and development (R&D) to improve understanding of, preparation for, and response to hazards. A comprehensive RAND analysis of current federal funding for research on hazard losses found that programs solely dedicated to hazard loss reduction receive the least funding, while work on weather hazards and broadly related research on climatology, atmospheric science, and oceanography receive the most. Much of this R&D spending supports short-term prediction capabilities, even though such measures have limited loss reduction potential. While prediction can generally move individuals out of harm's way, long-term loss reduction strategies could improve the resilience of communities and infrastructure, resulting in less property damage and reduced rebuilding costs. The study concludes that a comprehensive national loss database and greater use of loss modeling would assist in identifying essential research areas, and that R&D activities should be reoriented toward longer-term efforts, focusing on technologies that can reduce infrastructure losses and better protect individuals and property. Finally, the study offers recommendations on a more thoughtful framework for the role of research in hazard loss reduction.