The paper deals with the nature of religious knowledge, in particular the legitimacy of cognitive statements in religion. I suggest that neither science is as objective, nor religion as subjective as logical positivists use to argue. The starting point of my argument is Kuhn’s paradigmatic approach and Lakatos’s methodology of scientific research programs. These concepts indicate that scientific models are products of a creative analogical imagination; data are theory-laden; comprehensive theories are resistant to falsification; and strict criteria for paradigm choice can hardly (if at all) be found. Actually, all of these subjective features are more prominent in religion. Nevertheless, on my view, the difference between science and religion is a question of degree rather than absolute contrast.