The Southern Ocean (SO) is among the regions on Earth that are undergoing regionally the fastest environmental changes. The unique ecological features of its marine life make it particularly vulnerable to the multiple effects of climate change. A network of Marine Protected Areas (MPAs) has started to be implemented in the SO to protect marine ecosystems. However, considering future predictions of the Intergovernmental Panel on Climate Change (IPCC), the relevance of current, static, MPAs may be questioned under future scenarios. In this context, the ecoregionalization approach can prove promising in identifying well‐delimited regions of common species composition and environmental settings. These so‐called ecoregions are expected to show similar biotic responses to environmental changes and can be used to define priority areas for the designation of new MPAs and the update of their current delimitation. In the present work, a benthic ecoregionalization of the entire SO is proposed for the first time based on abiotic environmental parameters and the distribution of echinoid fauna, a diversified and common member of Antarctic benthic ecosystems. A novel two‐step approach was developed combining species distribution modelling with Random Forest and Gaussian Mixture modelling from species probabilities to define current ecoregions and predict future ecoregions under IPCC scenarios RCP 4.5 and 8.5. The ecological representativity of current and proposed MPAs of the SO are discussed with regards to the modeled benthic ecoregions. Twelve benthic ecoregions were determined under Present conditions, they are representative of major biogeographic patterns already described. Our results show that the most dramatic changes can be expected along the Antarctic Peninsula, in East Antarctica and the sub‐Antarctic islands under both IPCC scenarios. Our results advocate for a dynamic definition of MPAs, they also argue for improving the representativity of Antarctic ecoregions in proposed MPAs and support current proposals of CCAMLR for the creation of Antarctic MPAs.