In March 1972, Stephen Hanauer, a technical advisor with the Atomic Energy Commission, met with Norman Rasmussen, a nuclear engineering professor at the Massachusetts Institute of Technology. The AEC had recruited Rasmussen to develop a report, The Reactor Safety Study (WASH-1400), to estimate the probabilities and consequences of a major nuclear power plant accident. With thousands of safety components in a modern reactor, the task was mind-boggling. Rasmussen proposed a novel approach based on more powerful computers, “fault tree” methodology, and an expanding body of operational data. By calculating and aggregating probabilities for innumerable failure chains of components, he believed he could develop a meaningful estimate of overall accident risk. WASH-1400 would be a first-of-its-kind probabilistic risk assessment (PRA).