Patterns of seabird species' distributions differ between the Antarctic and the Arctic. In the Antarctic, distributions are annular or latitudinal, with strong similarities in species composition of seabird communities in all ocean basins at a given latitude. In the Arctic, communities are arranged meridionally, and show strong differences between ocean basins and, at a given latitude, between sides of ocean basins. These differences between the seabird communi ies in the Northern Hemisphere and the Southern Hemisphere reflect differences in the patterns of flow of major ocean current systems. At smaller spatial scales, in both hemispheres the species composition of seabird communities is sensitive to changes in watermass characteristics. The distribution of avian biomass is affected by both physical and biological features of the ocean. In the Antarctic, much seabird foraging is over deep water, and withinseason, small-scale patchiness in prey abundance and availability in ice-free waters is likely to be controlledprimarily by the behavior of the prey, rather than by physical features. Thus, prey availability may be unpredictable in time and space. In contrast, in the Northern Hemisphere, most seabirdforaging is concentrated over shallow continental shelves, where currents interact with bathymetry to produce predictable physical features capable of concentrating prey or making prey more easily harvested by seabirds. Ice cover appears to be the most important physical feature in the Antarctic. An entire community of birds is specialized to use prey taken near the ice edge. These prey consist of a variety of species, some of which are normally found much deeper in the water than the birds takingthem can dive. The open-water portion of the marginal ice zone is also an important foraging habitat for Antarctic marine birds. In the Arctic, a food web based on underice algae is used by marine birds, but few if any data exist on avian use of the open water segment of the marginal ice zone. Recent simultaneous surveys of birds and their prey indicate that only rarely does the small-scale abundance of birds match that of their prey; correlations between predators and prey aregenerally stronger at larger scales. Evidence is accumulating in the Antarctic that the largestaggregations of krill may be disproportionately important to foraging seabirds.