CCAMLR aims to develop a feedback approach to aid ecosystem-based management (EBM) of Antarctic krill fisheries. It is important to assess whether a feedback approach is likely to achieve the multiple objectives that EBM implies in the complex and uncertain conditions typical of Antarctic marine ecosystems. This study used Model Predictive Control (MPC) to achieve objectives for a harvested species, its predators and the fishery, in a simulation model that incorporates uncertainty and spatial and trophic complexity. The approach adjusted spatially resolved annual catch limits in response to estimates of the state of the system. It suggests that feedback management is both feasible and a more effective way to achieve multiple objectives than fixed catch limits, which are currently used to manage Antarctic krill fisheries. The study demonstrates that optimisationbased approaches such as MPC are computationally capable of dealing with EBM-type problems. They are also useful for assessing the feasibility of candidate management policies or objectives, and characterising the trade-offs that they imply. This study characterises the trade-off between catch levels and the risk of harvested species biomass falling to unacceptable levels.