1. We used the CCAMLR 2000 synoptic survey acoustic dataset to compare pelagic and shelf SSMUs in terms of the frequency of 1nm integration units with krill densities above and below a range of threshold values. 2. In general pelagic SSMUs had markedly lower frequencies of fishable integration units where the threshold density for fishing was set at 100 gm2 following Kasatkina (2003). 3. This was also true with other threshold values in the range 25 to 200 gm2 although the difference was smaller for higher thresholds. 4. The contrasting probability of encountering fishable concentrations in pelagic and shelf SSMUs suggests that the fishery will be less efficient and, perhaps, less economically viable in pelagic SSMUs. This is despite the fact that the absolute abundance of both krill and fishable concentrations of krill is higher in pelagic than in shelf SSMUs. 5. We also derived empirical relationships between SSMU-scale krill density and the frequency of fishable integration units for the full range of threshold levels. These relationships will be useful for linking the scale represented in operating models with those that affect the behaviour and performance of the fishery.