The importance of geographic isolation in larval availability vs. postsettlement mortality as factors influencing coral community structure was examined. Coral recruitment and mortality patterns were assessed on the Great Barrier Reef (Australia) via settling plates implanted on three reefs located 16, 58, and 117 km from shore (across the continental shell) at depths of 3 and 15 m. In a controlled experiment, plates were transplanted after 6 months of exposure to other reefs, covering all possible between-reef combinations. Recruitment was higher on the outer half of the shelf and in deeper water. Acropora and Seriatopora were the major recruits on the midand outershelf reefs, where adult Acropora dominates. Porites and Alveopora dominated the inner-shelf reef, where Porites is a dominant adult. Mortality was lowest on the midshelf reef. Inshore recruits suffered greater losses in competition for space. Coral larvae settled cryptically, except in deeper inshore waters (low light). Taxonomic gradients occurred across the shelf, suggesting some geographic isolation, and implying that local larval pools differ. Spat transplanted from the midshelf reef to inshore or offshore reefs suffered higher mortality rates. Transplants from offshore survived better on the midsheff reef. Transplants from inshore to midor offshore reefs appeared to survive well. Larval availability, regional seeding of reefs, and postsettlement mortality may together play important roles in influencing coral community structure. Two sets of primary factors are known to influence the structure of ecological communities in their earliest stages of development and their subsequent direction of succession (Levinton 1982). The first is dispersal (Scheltema 197 1,19 86) and the probability of colonization of a given habitat (Sale 1978; Kubitzki 1983; Gaines et al. 1985). The second is postsettlement selection (Jackson 1977; Levinton and Lassen 1978; I Present address: Resource Assessment Commission, Locked Bag No. 1, Queen Victoria Terrace, Canberra, A.C.T. 2600, Australia. Acknowledgments I thank J. S. Bunt for providing the impetus and resources to conduct this study. I also thank J. H. Carleton for assistance in the field and laboratory throughout the study and for assistance with data analyses. H. Crenshaw and numerous other volunteers also assisted in the field. G. Russ, J. Pandolfi, R. Reichelt, and D. Williams provided comments on the manuscript. I am grateful to the Natural Sciences and Engineering Research Council (NSERC) of Canada, the Department of Industry, Technology and Commerce (DITAC) of Australia, and the Department of Geology at McMaster University (Canada) for the opportunity to complete this manuscript. This study was supported by the Australian Institute of Marine Science. Australian institute of Marine Science Contribution 523. Sammarco 1980, 1982). The question of which of these processes has greater influence has received a reasonable amount of attention of late, particularly in the marine environment (Doherty and Williams 1988; Thresher and Brothers 1989; Lewin 1986). Several taxonomic groups exhibit broadscale patterns of zonation across the continental shelf in the central region of the Great Barrier Reef. Here, the shelf is 120 km wide, and a gradient of community structure has been documented for such groups as scleractinian corals (Done 1982, 1983), alcyonacean soft corals (Dinesen 1983), phytoplankton (Revelante et al. 1982), zooplankton (Ikeda et al. 1980; Sammarco and Crenshaw 1984) algae (Drew 1983), fish (Williams and Hatcher 1983), and sponges (Wilkinson and Trott 1985; Reichelt et al. 1986). With respect to corals, Acropora is a commonly occurring genus on the Great Barrier Reef that can dominate reefs in shallow water (Wallace 1978; Veron and Wallace 1984). Done (1982) in an extensive study of coral communities in the central region of the Great Barrier Reef, labeled the midand outer-shelf reefs “Acropora reefs.” EIe contrasted these against inshore ones, which he