Aim: For decades, biogeographers have used patterns of animal and plant diversity to identify areas that could be considered for conservation efforts. Since richness is not the best framework to conserve biodiversity, ecologists have been using patterns of complementarity to ensure that sites selected for inclusion in a reserve network complement those already selected. They also have investigated their association with the environment to conserve biodiversity in terrestrial realms successfully. In this study, we extended these ideas to marine realms. Specifically, we investigated if complementarity patterns vary with latitude, and what are the reasons for the observed patterns. Location: Global. Time period: Present day. Major taxa studied: Marine mammals. Methods: Global distribution maps for 123 marine mammals were obtained from the International Union for the Conservation of Nature (IUCN) Spatial Database to determine the global presence/absence of species. Random forest models were used to investigate the relationship between oceanographic variables and patterns of complementarity-based marine mammal site importance, calculated using the software Zonation. Results: Complementarity maps for all marine mammal species show a gradient of increasing importance from pelagic towards coastal areas, with patches of high priority in the Southern, North Pacific, and Arctic Oceans. Conversely, maps depicting hotspots of richness (HRs) of marine mammals show that HRs are highly biased towards tropical zones. Random forests identified temperature, bathymetry and salinity as the critical drivers of high conservation priority for marine mammals at a global extent. Main conclusions: Our results support the tenet that site complementarity can be modelled and predicted as a function of environmental variables. Because marine mammals face a higher level of threats compared to their land counterparts, our results can help stakeholders and citizens to advocate for actions in priority areas. Also, complementarity patterns could provide important insight into how patterns might change in the future, given the impacts of climate.