The measurement of spatial overlap between predators and fisheries exploiting a common prey source is dependent upon the measurement scale used; inappropriate scales may produce misleading results. Previous assessments of the level of overlap between predators and fisheries for Antarctic krill (Euphausia superba) in the region of the South Shetland Islands used different measurement scales and arrived at contradictory conclusions. At-sea data from observations of krill predators during the CCAMLR 2000 Survey were used to identify the areas of potential overlap with fisheries in the Scotia Sea and to determine the scale at which such overlap should be measured. The relationship between autocorrelation and sampling distance was used to identify the characteristic scales of the distribution of predators, krill and krill fisheries, and an effort-corrected index of relative abundance as a function of distance from land was used to identify the characteristics of areas of high potential for overlap. Despite distinct differences in foraging ecology, a group of krill-dependent species including chinstrap penguin (Pygoscelis antarctica), (Antarctic) fur seal (Arctocephalus sp. (gazella)) and white-chinned petrel (Procellaria aequinoctialis) showed similar patterns of distribution; the relative abundances were highest at 60-120 km from land and decreased sharply at distances greater than 150 km from land. There were more inter-specific differences in the characteristic scales, which were of the order of 50-100 km. Antarctic krill had a characteristic scale of approximately 200 km and the relationship with distance from land showed a log-linear decline. Krill fisheries operate at a scale of 150 km and occur almost entirely within 100 km of land. The requirement of land for breeding and the biological and oceanographic conditions that produce the high concentrations of krill associated with those land areas produce a system in which the demand for Antarctic krill from fisheries and predators is essentially co-extensive. The areas of greatest potential overlap are within 150-200 km of land and to accommodate the scales of operation of the processes involved the extent of such overlap in these areas should be assessed at scales of 70-100 km.