In recent years, the question of naturalism in the study of religions has been increasingly debated. Primarily, these discussions converge in the widely held view that naturalism is the only way for religious studies as an academic enterprise to exclude supernaturalist assumptions from its methodology. While I fully agree with this view, I argue that naturalism is usually formulated with the help of metaphysical assumptions, which are problematically embodied in the location problem, that is, the problem of how to locate certain phenomena, such as meanings and values, in the order of nature. By unfolding the dynamic between the elements of the location problem, I show that the kind of naturalism based on Wittgenstein’s thought prevents the location problem from arising and can serve as a balanced version of naturalism for use in the study of religion. While metaphysical naturalism often leads to dilemmas, within Wittgenstein’s kind of naturalism, it seems possible both to maintain anti-supernaturalism in the study of religion and to resist the metaphysical temptations hidden in our assumptions about language. These two features make Wittgenstein’s naturalism truly methodological.