Description: Landscape maps based on multivariate cluster analyses provide an objective and comprehensive view on the (marine) environment. They can hence support decision making regarding sustainable ocean resource handling and protection schemes. Across a large number of scales, input parameters and classification methods, numerous studies categorize the ocean into seascapes, hydro-morphological provinces or clusters. Many of them are regional, however, while only a few are on a basin scale. This study presents an automated cluster analysis of the entire Atlantic seafloor environment, based on eight global datasets and their derivatives: Bathymetry, slope, terrain ruggedness index, topographic position index, sediment thickness, POC flux, salinity, dissolved oxygen, temperature, current velocity, and phytoplankton abundance in surface waters along with seasonal variabilities. As a result, we obtained nine seabed areas (SBAs) that portray the Atlantic seafloor. Some SBAs have a clear geological and geomorphological nature, while others are defined by a mixture of terrain and water body characteristics. The majority of the SBAs, especially those covering the deep ocean areas, are coherent and show little seasonal and hydrographic variation, whereas other, nearshore SBAs, are smaller sized and dominated by high seasonal changes. To demonstrate the potential use of the marine landscape map for marine spatial planning purposes, we mapped out local SBA diversity using the patch richness index developed in landscape ecology. It identifies areas of high landscape diversity, and is a practical way of defining potential areas of interest, e.g. for designation as protected areas, or for further research. Clustering probabilities are highest (100%) in the center of SBA patches and decrease towards the edges (〈 98%). On the SBA point cloud which was reduced for probabilities 〈98%, we ran a diversity analysis to identify and highlight regions that have a high number of different SBAs per area, indicating the use of such analyses to automatically find potentially delicate areas. We found that some of the highlights are already within existing EBSAs, but the majority is yet unexplored.