Poster presentation at ATLAS 3rd General Assembly. An objective statistical approach has been used to hierarchically partition the benthic area within the Azores EEZ into environmental units of similar characteristics. High-dimensional climatology and geomorphology data were projected and clustered onto a two-dimensional space following a three-step method: 1) principal component analysis (PCA), 2) K-means clustering and 3) optimal cluster determination. A preliminary multiple-scale analysis was also considered to better detect regional processes and structures. On the one hand, this classification can help defining sampling strategies accounting for environmental representativity and uniqueness at different scales and assessing the value of environmental proxies for predicting biological turnover. On the other hand, it can help structuring marine spatial planning and conservation strategies in areas where biological data is scarce.