The various Sraffa-based criticisms of marginalist theory which emerged in the 1960s and 1970s took on many different manifestations-reswitching, capital-reversing, upward sloping relative demand curves for primary inputs, etc., etc. Yet behind these varied forms there always stood one and the same fundamental fact, namely that the relative prices of produced commodities changed (in general) whenever the uniform rate of profit changed, whether or not the technique of production also changed. That fundamental fact also stood, of course, at the root of the notorious 'transformation problem' of Marxian economics. Since Geoff Harcourt has been greatly interested in such questions, it may not be inappropriate to devote this contribution to his Festschrift to a consideration of how much relative prices can change as the rate of profit changes. It is very well known that only under very special conditions do they not vary at all. But when they do vary, by how much do they vary? This question has received some attention in the context of empirical, input-output studies (see, for example, Bienenfeld (1988) and Ochoa (1989)) but it would, presumably, be desirable to see whether any limits to relative price movement can be established theoretically. For while it is certainly true that any degree of relative price sensitivity to the rate of profit is sufficient to undermine many well-known constructions in economic theory, the extent of such sensitivity is still of importance in determining, for example, the probability of reswitching, the likely magnitude of the errors involved in approximating production prices by 'Marxian values', etc. It should perhaps be said at once that what follows does not provide definitive answers to the questions which are implicitly being raised here but it may be hoped that it does provide some suggestions as to how such questions might be dealt with.