Disasters are typically unforeseen, causing most social and behavioral studies about disasters to be reactive. Occasionally, predisaster data are available, for example, when disasters happen while a study is already in progress or where data collected for other purposes already exist, but planned pre-post designs are all but nonexistent. This gap fundamentally limits the quantification of disasters’ human toll. Anticipating, responding to, and managing public reactions require a means of tracking and understanding those reactions, collected using rigorous scientific methods. Oftentimes, self-reports from the public are the best or only source of information, such as perceived risk, behavioral intentions, and social learning. Significant advancement in disaster research, to best inform practice and policy, requires well-designed surveys with large probability-based samples and longitudinal assessment of individuals across the life-cycle of a disaster and across multiple disasters.