This article deals with the issue of using concepts and categories in social sciences with particular focus on the concept of "religion". This concept has been critically analyzed by several authors, e.g. by Timothy Fitzgerald, Russell T. McCutcheon and Daniel Dubuisson. The category of religion is seen by its critics as a biased and manipulative socio-cultural construct, originating in the specific social and cultural environment of Western (Euro-American, Christianity-based) societies. However, the critique of the notion of religion as a socio-culturally constructed concept does not necessarily mean that the concept itself should be abandoned or that it is of no practical use for the description and analysis of non-Western social and cultural phenomena. This contention is evidenced with examples from the Japanese context. The notion of religion (shūkyō) cannot easily be proclaimed to be a Western import into Japan; it has older roots in Japanese history. The extent and content of this concept, as well as its place in a wider conceptual network, have certainly been transformed in the process of cultural exchange with different Euro-American environments but this does not imply that this concept is simply a tool of colonial powers, nor does it preclude its analytical usefulness.