Acoustic estimates of the densities of Antarctic krill, Euphausia superba, in areas around South Georgia (SG) and Elephant Island (El) were compared for seven austral summers between 1981 and 1997. Estimated densities of krill at El were most often higher than at SG, although this may simply have been a function of differences in survey and data analysis techniques used at each site. More interestingly, the magnitudes of abundance and between-year gradients of change of abundance at each site were mirrored by those at the other location; for example, 1991 and 1994 were years of very low krill density at both SG and El. There was no apparent lag in changes in abundance at each site, and ranked between-year gradients of change in abundance at both locations were significantly correlated. These pronounced similarities suggests that densities of krill at both locations are directly linked, and may be impacted by the same gross physical and biological factors (e.g. recruitment, dispersal etc.), acting over the same temporal and spatial scales. The observed concordance also implies that the pelagic ecosystems at these widely separated sites (approximately 1500 km distant at opposite sides of the Scotia Sea) are not operating in isolation. Possible mechanisms linking krill population processes in the areas around South Georgia and Elephant Island are discussed.