It has sometimes been claimed that rudist bivalves competitively displaced corals from reef frameworks during the Cretaceous. This hypothesis combines two assertions: (1) that the autecology of rudists was convergent with that of reef-building corals; and (2) that rudist formations commonly developed as reefs. We dispute both assertions, and thus reject the hypothesis of competitive displacement. We argue instead that mobile sediments, rather than frameworks, dominated the margins and tops of most of the extensive carbonate platforms of the period, and that it was on these, non-reefal, biotopes that the rudists flourished. Definitions of reefs tend to combine two major elements: (1) a robust biogenic framework (with accompanying sedimentary and diagenetic components); and (2) topographical relief. Such definitions are clearly rooted in Recent coral reefs, in which endosymbiotic zooxanthellae permit the extensive growth of colonial coral frameworks in shallow but relatively nutrient-poor waters, coralline algae and cementation may contribute significantly to the growth of rigid structures, and topography is largely the legacy of Pleistocene changes in sea-level. In rudist formations, in contrast, individual rudist congregations are volumetrically limited, relative to sediment. They are often loosely constructed, and they evidently showed little, if any, original relief. Tabular and small lenticular units predominate. These differences in structure and palaeoenvironmental situation between rudist and modern coralgal associations reflect the different autecologies of the constituent organisms. The clonal growth of corals predisposes them to the development of frameworks projecting above the sediment surface (herein termed superstratal growth). By contrast, the aclonal development of rudists was better suited to the opportunistic occupation of a variety of temporarily available substrata, by large numbers of individuals. In particular, elevator rudists (in which the entire commissure exhibited upward growth) evidently grew with their shells largely embedded in, and supported by, the ambient sediment (herein termed constratal growth). Moreover, the tolerances and growth responses of rudists to such factors as water turbidity, nutrients and current regime were quite different from those of the majority of reef-building corals. Despite repeated assertions in the literature that rudists possessed zooxanthellae, only a few species show any evidence for such a symbiosis and other evidence suggests that most lacked them. In view of these differences in their preferred biotopes, competition between rudists and corals is doubtful, even though members of both groups co-occur in many areas. The relative decline of coral frameworks in the Cretaceous was thus probably independently caused. © 1995.