No matter one's wealth or social position, all are subject to the threats of natural hazards. Be it fire, flood, hurricane, earthquake, tornado, or drought, the reality of hazard risk is universal. In response, governments, non-profits, and the private sector all support research to study hazards. Each has a common end in mind: to increase the resilience of vulnerable communities. While this end goal is shared across hazards, the conception of how to get there can diverge considerably. The earthquake and hurricane research endeavors in the US provide an illustrative contrast. The earthquake community sets out to increase resilience through a research process that simultaneously promotes both high quality and usable - preparedness-focused - science. In order to do so, the logic suggests that research must be collaborative, responsive, and transparent. Hurricane research, by contrast, largely promotes high quality science - predictions - alone, and presumes that usability should flow from there. This process is not collaborative, responsive, or transparent. Experience suggests, however, that the latter model - hurricane research - does not prepare communities or decision makers to use the high quality science it has produced when a storm does hit. The predictions are good, but they are not used effectively. Earthquake research, on the other hand, is developed through a collaborative process that equips decision makers to know and use hazards research knowledge as soon as an earthquake hits. The contrast between the two fields suggests that earthquake research is more likely to meet the end goal of resilience than is hurricane research, and thus that communities might be more resilient to hurricanes were the model by which research is funded and conducted to change. The earthquake research experience can provide lessons for this shift. This paper employs the Public Value Mapping (PVM) framework to explore these two divergent public value logics, their end results, and opportunities for improvement. © 2011 Springer Science+Business Media B.V.