The data is the output of the global groundwater recharge (GG-R) model. The GG-R model is a grid-based model and is developed and implemented in Python to simulate the daily diffuse rain-fed global groundwater recharge. The model consists of three layers: topsoil (root zone), subsoil, and aquifer. The GG-R model calculates the exchange of water between topsoil and atmosphere, as well as surface runoff, topsoil recharge, water volume in soil layers, subsoil recharge from topsoil, capillary rise from the subsoil to the topsoil, and groundwater recharge, all on a daily time step and grid-based values. The model covers the spatial extent from 180.0°W to 180.0°E longitudes and 60.0°N to 60.0°S latitudes and a temporal range from January 2001 to December 2020 with a spatial resolution of 0.1°×0.1° and daily temporal resolution. The output provided here is the main result of the GG-R model and is the annual grid-based global groundwater recharge (R_gw) from 2001 to 2020 and 0.1°×0.1° spatial resolution.