In arid and semi-arid environments focussed infiltration of rain and snowmelt water under topographic depressions is an important mechanism of groundwater recharge. Quantifying the aggregated recharge from numerous small depressions is a major challenge in water resource management. Building on field-based investigations into the surface water-groundwater interaction of individual depressions and their catchments (i.e. uplands) in the Canadian Prairies, we have developed a simple water balance model to simulate groundwater recharge considering the hydrological coupling of a depression-upland system. The model is based on the Versatile Soil Moisture Budget (VSMB), which has been widely used in the Canadian Prairies to simulate soil moisture conditions. We evaluated the new model, VSMB Depression-Upland System (VSMB-DUS) using field data consisting of an artificial flooding experiment and long-term monitoring of a depression in Alberta, Canada. The model captured surface water level, soil moisture, and groundwater responses to the artificial flooding with a reasonable accuracy, and also represented the inter-annual variability of recharge fluxes during a five year period (2007-2011) including dry and wet years. Simulated annual recharge varied between 12 and 45 mm, whereas annual precipitation varied between 453 and 597 mm during the five-year period. The VSMB-DUS tends to over- or under-estimate snowmelt runoff in individual years, however, simulated recharge was only slightly impacted by the errors in snowmelt runoff estimation. Due to its computational efficiency and robust algorithms, the VSMB-DUS will provide a useful tool for estimating aggregated recharge in a large-scale model grid cell containing hundreds of depression-upland systems.