The term hermatypic, as widely used in the literature of extant and fossil Scleractinia, includes, by definition (Wells 1933), the confusing generalization of equating reef-building with containing zooxanthellae. In course of time the use of the term diverged into denoting organisms which are either reef-building (including calcareous Rhodophyta) or those that contain zooxanthellae (including soft Alcyonaria). The equation: reef-building corals harbour zooxanthellae and vice-versa, is invalidated by reviewing the various ecological categories of corals such as: reef-building species without the support of zooxanthellae, zooxanthellae-containing corals which inhabit but do not build reefs, zooxanthellae-containing, non-reef-building corals beyond the bathymetric and latitudinal limits of reefs, and framework-erecting corals in deep waters without zooxanthellae. Former attempts to improve the original definition of hermatypic are shown to be insufficient to match the ecological diversity of corals. A strict terminological separation of the properties zooxanthellae-containing, reef-building and (more generally) framework-building is provided by the use of the revised, respectively new terms zooxanthellate, hermatypic and constructional (with the respective antonyms azooxanthellate, ahermatypic and nonconstructional). This terminology also applies to non-scleractinians.