Climate warming is intensifying soil moisture droughts worldwide, with serious consequences for agriculture and terrestrial ecosystems. Robust assessment of their severity, extent, evolution and impacts on natural and managed systems requires globally consistent, harmonized and event-based datasets with sub-monthly temporal resolution. Here, we introduce a global catalogue of 755 Global Land soil moisture Drought Events (GLDEs) over the past four decades (1980–2024) derived from the gridded root-zone soil moisture estimates produced by two structurally different models: the hydrological model mHM and the soil water balance model SoilClim. Drought events are identified using a combination of 10th- and 20th-percentile soil moisture anomaly thresholds and clustered with a machine-learning-based OPTICS algorithm applied to 10-day averaged soil moisture fields. Each GLDE is characterized using seven severity categories (from extremely weak to extremely severe) and seven dynamic categories (from extremely static to extremely dynamic) based on four spatiotemporal and three motion-related characteristics. Extremely severe (7s) and extremely dynamic (7d) GLDEs occur most frequently in South America, followed by North America, whereas Eurasia experiences the longest and most extensive events. The dataset of all 755 GLDEs with their statistical characteristics is publicly available. Overall, the GLDE catalogue provides a unique, globally consistent basis for future investigations into the drivers and impacts of dynamically evolving drought events.