The 'functional region' typology divides the global cropland into clusters based on pre-selected attributes. These attributes stand for the agricultural functionality (yield and irrigation requirements) and environmental state of croplands (soil loss, aridity, and species loss). This approach allows identifying more suitable agricultural regions (e.g., higher yields and lower environmental impacts), or less suitable regions (e.g., water-stressed regions). The functional regions typology includes twelve clusters and is based on the self-organized maps algorithm (SOM). The functional regions typology and the associated distance file are included in this dataset.