Data from a probability sample of a state university faculty are analyzed to determine whether academic discipline is predictive of faculty religiosity. Wide variations observed on four of dock's proposed dimensions of religiosity are compared to the scientist-nonscientist dichotomy of fields and to “scholarly distance from religion,” a construct delineated in the paper. The scientist-nonscientist scheme is not predictive of religiosity scores. Scholarly distance from religion successfully orders the scores. A relationship between scholarly distance from religion and religiosity seems to be interpreted partly by the effect of scholarly distance from religion on cognitive differentiation of religion and on social support for religiosity, especially the former. Selectivity of discipline and present religiosity appear to be based on childhood religious background, but controlling for childhood religiosity does not alter the original relationships greatly.