Fermi knew that when an atom splits it releases other neutrons, and he was quick to realize that under the right conditions those neutrons could go on to split other atoms in a chain reaction. This would lead to one of two things: a steady generation of energy in the form of heat or a huge explosion. If each splitting atom caused one released neutron to split another atom, the chain reaction was said to be "critical" and would create a steady release of heat energy. But if each fission event released two, three, or more neutrons that went on to split other atoms, the chain reaction was deemed "supercritical" and would rapidly cascade into an almost instantaneous, massive, explosive release of energy—a bomb. In the climate of the times, with the world on the brink of war, there was little doubt in which direction the main research effort would turn. Fermi, who had emigrated to the United States, became part of the top-secret American effort known as the Manhattan Project, which, in an astonishingly short period of time from its beginnings in 1942, turned fission's potential into the reality of the world's first atomic bombs.

The Manhattan Project, headed by General Leslie Groves of the Army Corps of Engineers, included experimental facilities and manufacturing plants in several states, from Tennessee to Washington. Dozens of top-ranking physicists and engineers took part. One of the most significant breakthroughs was achieved by Fermi himself, who in 1942 created the first controlled, self-sustaining nuclear chain reaction in a squash court beneath the stands of the University of Chicago stadium. To do it, he had built the world's first nuclear reactor, an achievement that would ultimately lead to the technology that now supplies a significant proportion of the world's energy. But it was also the first practical step toward creating a bomb.

Fermi recognized that the key to both critical and supercritical chain reactions was the fissionable fuel source. Only two potential fuels were known: uranium-235 and what was at the time still a hypothetical isotope, plutonium-239. (An isotope is a form of a given element with a different number of neutrons. The number refers to the combined total of protons and neutrons in the nucleus.) Uranium-235 exists in only 0.7 percent of natural uranium ore; the other 99.3 percent is uranium-238, a more stable isotope that tends to absorb neutrons rather than split and that can keep chain reactions from even reaching the critical stage. Plutonium-239 is created when an atom of uranium-238 absorbs a single neutron.