The magnitude of flood damage in the United States, combined with the uncertainty in current estimates of flood risk, suggest that society could benefit from improved scientific information about flood risk. To help address this perceived need, a group of researchers initiated an interdisciplinary study of climate variability, scientific uncertainty, and hydrometeorological information for flood-risk decision making, focused on Colorado's Rocky Mountain Front Range urban corridor. We began by investigating scientific research directions that were likely to benefit flood-risk estimation and management, through consultation with climatologists, hydrologists, engineers, and planners. In doing so, we identified several challenges involved in generating new scientific information to aid flood management in the presence of significant scientific and societal uncertainty. This essay presents lessons learned from this study, along with our observations on the complex interactions among scientific information, uncertainty, and societal decision making. It closes by proposing a modification to the “end to end” approach to conducting societally relevant scientific research. Although we illustrate points using examples from flood management, the concepts may be applicable to other arenas, such as global climate change.