History offers many examples of the recurring tensions between science and organized religion, but as part of the paper’s motivating evidence we also uncover a new fact: in both international and cross-state U.S. data, there is a significant and robust negative relationship between religiosity and patents per capita. Three long-term outcomes emerge. First, a "Secularization" or "Western-European" regime with declining religiosity, unimpeded science, a passive Church and high levels of taxes and transfers. Second, a "Theocratic" regime with knowledge stagnation, extreme religiosity with no modernization effort, and high public spending on religious public goods. In-between is a third, "American" regime that generally (not always) combines scientific progress and stable religiosity within a range where religious institutions engage in doctrinal adaptation.