Professor Kai Nielsen is one of the most forceful proponents of the view that theological assertions have no factual reference because they are compatible with any empirical state of affairs; no evidence, it is alleged, is allowed to count as falsification of such assertions, and therefore they spuriously purport to be what they are not. In this he follows the well-known essay by Professor Antony Flew in which the same argument was advanced, and Nielsen's own most recent contribution on the subject in Religious Studies should be read against the background of his earlier controversy with Professors John Hick and George Mavrodes in The Canadian Journal of Theology. In this latest article he broadens the canvas to question the significance of applying language and concepts appropriate to the realm of empirical observation to anything that is held to transcend it; and this is the substance of Part II of his paper, in which he engages in debate with Professor Ian Crombie, treading well-worn ground that has been the subject of innumerable philosophical discussions. The rest of the paper, however, is a restatement of his earlier contention, and it is with this more limited issue that I shall be concerned in what follows. For I believe that Nielsen is wrong in two quite fundamental respects: (1) he is mistaken in supposing that theological assertions are compatible with any empirical state of affairs: (2) he bases his critique of theological assertions on a far too uncritical appraisal of the principles of verification and falsification in the case of ordinary factual statements. If I am right, the larger questions raised in Part II will still remain open, but it is high time that the ‘verification-falsification argument’ should be shown to be unfounded.