Studies of the benthos between 30 and 60 m at Cape Armitage, McMurdo Sound. Antarctica, reveal an epifaunal community in which sponges and their asteroid and nudibranch predators predominate. Field experiments demonstrated that, with the exception of Mycale accrata, the growth rates of the sponges are too slow to measure in one year. Mycale, however, was observed to increase its mass as much as 67%. Because of its more rapid growth rate, Mycale appears to be the potential dominant in competition for substratum space, the resource potentially limiting to the sessile species. This conclusion is supported by observations of Mycale growing over and, in some cases, apparently having smothered many other sessile species representing at least three phyla. The densities and size frequency distributions of all the predators were measured; numerous feeding observations allowed an accurate appraisal of dietary compositions. Because of the predators' very slow consumption rates, however, direct measures of ingestion and its impact on prey populations were not possible. Estimates of the ingestion rates were derived from measurements of predator respiration rates, growth rates, and gonad growth. Data from the field surveys and the energetics studies suggest that Mycale is prevented from dominating the space resource by the predation of two asteroids. Perknaster fuscus antarcticus and Acodontaster conspicuus. Adult Perknaster specialize on Mycale, and the sponge provides a small proportion of the diet of A. conspicuus. Acodontaster conspicuus and the dorid nudibranch Austrodoris memurdensis are the most important predators on three species of rossellid sponges (Rossella racovitzae, R. nuda, and Scolymastra joubini). Despite this relatively heavy consumption and despite the fact that none of these sponges has a refuge in growth from potential mortality from A. conspicuus, very large standing crops of the rossellid sponges have accumulated. This accumulation appears to result from predation on larval and young A. conspicuus and Austrodoris by Odontaster validus, which is primarily a detrital feeder and apparently acts as a filter against the settlement and survival of the A. conspicuus and Austrodoris larvae. In addition, predation upon adult A. conspicuus by O. validus and the actinian urticinopsis antarcticus annually kills approximately 3.5% of the A. conspicuus population. This mortality exceeds the apparent rate at which A. conspicuus escape the larval filter.