Scott Atran and colleagues say program flaws have hindered US counter-terrorism efforts. They propose an alternative design that links policy to “theoretically informed field research,” including “field data, collected in scientifically reliable ways, from terrorists, supporters and host populations.”