The mesopelagic zone (200 – 1000 m) may contain over 10 billion tonnes of fish, which is a potentially important future fishery resource. The major constituents of the mesopelagic fish community are myctophids (c. 250 species): although not fit for human consumption, they could be used as food for farmed animals and fish, and also as fertiliser for crops. During the daytime, myctophids form deep (300 – 1000 m), horizontally extensive (1000s of km) layers that are readily detectable with echosounders. The layers are known as Deep Scattering Layers (DSLs) because they scatter sound. At night, a proportion of fish in DSLs migrate upwards to access their prey (including zooplankton), and feed nearer the surface. The fish migrate to depth at dawn, and this 2-way vertical migration is an important process for vertical transport of nutrients and carbon in to the ocean interior. The depth and biomass of the mesopelagic community is influenced by environmental factors, giving rise to geographical variation. Akin to the surface biogeography described by A. Longhurst, which was based on primary production and regional oceanography, here we describe a mesopelagic biogeography, based on regional estimates of myctophid biomass and DSL depth, from a global dataset of echosounder observations. We estimate regional values of trophic efficiency and predict, using output from a coupled climate-ecosystem model (NEMO-MEDUSA-2.0), how these values may change by 2100. Ultimately, we provide a method for partitioning the ocean, across a range of scales, into distinct compartments/habitats, which are useful for mesopelagic study. In doing so, we highlight the differences in structure between the surface and mid-water communities and provide an informative map for both conservation and fisheries management.