An obscure paper about an earthquake in 869 C.E. that destroyed a castle town in northeastern Japan and a subsequent tsunami that inundated the surrounding area is now at the center of a growing debate about how quickly scientific findings can and should influence disaster-mitigation policies. A few years before the magnitude-9.0 Tohoku earthquake struck northeastern Japan on 11 March, a scientific consensus had begun to coalesce around the idea that a Jogan-like event could happen again. But that consensus did not influence seismic risk assessments, tsunami preparedness, or a review of the hardiness of the Fukushima Daiichi nuclear power plant. The need to revise earthquake probability analyses extends far beyond Japan.