Increased awareness about the present status of marine systems, including those where fisheries target the prey of natural predators, has led to recommendations about how information on the status of marine predators might be used to inform ecosystem-based management approaches. In the Antarctic, sustained ecological research has generated long-term data on the performance of marine predators, coupled with data that describe the density of their principal prey, Antarctic krill. Here we explore some of the longest time series of such data yet available, using records that were closely matched with consistent ecological time periods. Our results show that: (i) at the larger scale, some predator variables are correlated across sites and they therefore probably reflect common regional ecological conditions; (ii) at smaller scales, other variables depend upon local biological conditions, requiring that managers must evaluate the applicable scale of these predator indices if they are to be used in management decisions; (iii) variability in the krill population is evident at different spatial and temporal scales, some of which is important for predators, indeed when krill density is low, spatial variability and patchiness is probably a more important determinant of foraging success than density per se; (iv) previously documented relationships between predators and krill density are not apparent, suggesting that correlative relationships may break down as data series change in duration. We note that relationships that describe critical density thresholds for prey might not provide adequate management information, particularly at low krill densities. We suggest that a programme of work is needed to better characterise predator requirements, if ecosystem-based approaches are to be reliably implemented.