This review critically examines the literature on using satellite data for mapping landslides, focusing on the period from 1996 to 2022. The study created and analyzed an extensive bibliographic database from Web of Science, comprising 291 outputs from over 1,000 authors who studied nearly 700,000 landslides across all continents, with China and Italy leading in the number of studies. The analysis reveals that satellite imagery has been primarily used to detect and map two main types of landslides: flows and slides, often triggered by rainfall, earthquakes, or both. Manual detection was predominantly used until 2020, when artificial intelligence (AI) began to dominate. Despite significant progress, challenges remain in effectively and operationally using Earth Observation (EO) images for landslide detection and mapping. The review highlights the increasing availability and accessibility of EO data, driven by missions like Landsat and Sentinel, and discusses the future potential of EO in landslide mapping, including the integration of AI and the development of common guidelines for operational applications. The study also addresses the spatial and temporal distribution of research, the types of landslides analyzed, and the techniques used, emphasizing the role of AI in improving mapping accuracy. Finally, it suggests areas for future research to enhance the effectiveness of EO in landslide mapping and management.This review critically examines the literature on using satellite data for mapping landslides, focusing on the period from 1996 to 2022. The study created and analyzed an extensive bibliographic database from Web of Science, comprising 291 outputs from over 1,000 authors who studied nearly 700,000 landslides across all continents, with China and Italy leading in the number of studies. The analysis reveals that satellite imagery has been primarily used to detect and map two main types of landslides: flows and slides, often triggered by rainfall, earthquakes, or both. Manual detection was predominantly used until 2020, when artificial intelligence (AI) began to dominate. Despite significant progress, challenges remain in effectively and operationally using Earth Observation (EO) images for landslide detection and mapping. The review highlights the increasing availability and accessibility of EO data, driven by missions like Landsat and Sentinel, and discusses the future potential of EO in landslide mapping, including the integration of AI and the development of common guidelines for operational applications. The study also addresses the spatial and temporal distribution of research, the types of landslides analyzed, and the techniques used, emphasizing the role of AI in improving mapping accuracy. Finally, it suggests areas for future research to enhance the effectiveness of EO in landslide mapping and management.