<p>The various water fluxes of the global terrestrial water cycle are integral to the Earth system and the well-being of societies. However, fluxes occurring below the land surface, such as groundwater recharge and discharge, are more poorly constrained observationally than surface fluxes such as streamflow. Consequently, the broader relevance of these hidden fluxes is less well understood, and their global estimates are more uncertain. Here we combine multiple observational datasets and theoretical considerations within a Budyko-type water balance framework, providing a starting point for an enhanced understanding of subsurface water partitioning at large scales. Observations indicate that climatic aridity substantially influences subsurface flux partitioning, but that there is considerable variability in the need for further explanation. We show how this framework can be used to integrate empirical data, theoretical constraints and model-based insights to better understand subsurface flux partitioning and its controlling factors. Such a holistic approach is essential to better understand subsurface water cycling, especially in the face of increasing resource demands and climate change.</p>