Three hundred and four scleractinian colonies representing 58 species and 11 families were collected from six intertidal sites at Heron Island Reef, Great Barrier Reef. Tissue obtained from each colony was extracted with water, aqueous ethanol or a methanol-dichloromethane mixture. Extracts were assessed using a broad spectrum bioassay screen, incorporating micro-organisms, cells, invertebrates and vertebrates.At least one extract from each of 56 species (98%) exhibited activity against at least one bioassay regime. This is the first report of bioactivity in extracts of 46 species of coral and the first report of bioactivity of extracts of corals from the families Mussidae, Merulinidae, Siderastreidae, Oculinidae and Dendrophylliidae. Extracts from 56 species (98% of the 58 assayed) were toxic to at least one of the six metazoan organisms assayed. Cytolytic activity was present in extracts from 50 species (88% of the 57 tested), while extracts from 42 species (76% of the 55 species) possessed antimicrobial activity. Extracts from 31 species (62% of the 50 assayed) were toxic to cilia and/or flagella.There was marked variation in bioactivity between corresponding extracts from different colonies of a species. Variable results on three bioassays, namely toxicity to mice, cytolytic activity and antibacterial activity, were analysed by the Generalised Linear Interactive Modelling system (GLIM). Toxicity to mice, cytolytic and antibacterial activity are not significantly correlated with each other. The incidence of bioactivity in all bioassay regimes showed evidence of temporal variation. No model could be generated to explain satisfactorily the variation in the incidence of cytolytic nor of antibacterial activity. The model predicts that toxicity to mice will be high in extracts collected when average maximum monthly air temperature is high. Bioactivity in scleractinians appears to be idiosyncratic to each colony.Single colonies of the corals Lobophyllia corymbosa, Favites abdita, Favia matthaii, Favia stelligera, Platygyra daedalea, Leptoria phrygia, Cyphastrea serailia, Hydnophora exesa and Astreopora myriophthalma were permanently marked with bouys. Portions of colonies were removed up to seven times at intervals of two or three months. Aqueous extracts of the colony portions were assayed using six bioassay regimes namely, toxicity to mice, toxicity to a coral and a hydroid, cytolytic activity on sheep erythrocytes and sea urchin ova and for antimicrobial activity on eight bacterial species. The incidence of one type of bioactivity in an extract was not correlated with the incidence of any other type of activity in that extract. Although each coral colony provided extracts that affected at least two of the bioassay systems, different activity profiles were obtained from successive extracts of each colony. Thus there is a temporal component to the variable nature of bioactivity within a given colony of scleractinian coral.The potential of scleractinian toxins as mediators of ecological interactions was evaluated. Ten colonies of Goniopora tenuidens were placed individually in sea water for one hour after which time the sea water was assayed for toxicity to another hard coral, Galaxea fascicularis and to other colonies of G. tenuidens. Sea water from eight of the 10 colonies assayed was toxic to at least one of the coral species. Thus G. tenuidens can exude biologically active substances that adversely affect potential competitors, both interspecific and conspecific.It is concluded that bioactivity is widespread across scleractinian taxa and that scleractinian toxins are active against potential predators, competitors and fouling organisms. Because of intraspecific and intracolonial variation in bioactivity, toxicity of scleractinian corals is variable rather than a characteristic that can be defined for any species.