• We provide a method of analysing penguin tracking data to identify marine IBAs, based on pre-existing approaches but revised according to the specific ecology of Pygoscelis penguin species;

• Key improvements to the pre-existing approach include 1) refining track interpolation methods, and 2) revision of parameters for the kernel analysis (the smoothing factor and the utilization distribution);

• We applied the revised method to 24 datasets of tracking data (three species, seven colonies and three different breeding stages – incubation, brood and crèche);

• We compared the results of the marine IBA protocol with the results of habitat models, developed with data collected at other colonies;

• We show, for the first time, that maps of predicted distributions based on habitat suitability models can be used with a high degree of confidence to identify marine IBAs for penguins;

• The results reported here provide a new method for the designation of a network of marine IBAs in Antarctic waters for penguins, based on tracking data (when available) and habitat modelling, which can contribute to an evidence-based, precautionary, management framework for krill fisheries.