Groundwater, a key freshwater source in arid and semi-arid regions, faces growing stress from climate variability and human pressures, yet limited monitoring data constrain effective management. This study investigates the spatial and temporal variations of groundwater storage (GWS) across Afghanistan using data from the Gravity Recovery and Climate Experiment (GRACE), its successor GRACE Follow-On (GFO), and the Global Land Data Assimilation System (GLDAS) and to assess the influence of key climatic parameters on GWS dynamics. Results indicate that from 2004 to 2012, GWS remained relatively stable; however, from 2012 onward, a significant and accelerating decline occurred, with total loss reaching approximately - 13 cm by 2024. The Mann-Kendall test confirmed a significant decreasing trend (Z = - 10.16, p < 0.05). Correlation and regression analyses revealed that snowmelt (r = 0.56) and precipitation (r = 0.45) are the dominant drivers of groundwater recharge, while land surface temperature (r = - 0.09) has a weaker, negative influence due to enhanced evapotranspiration. Lagged correlation analysis identified the strongest positive correlation (r = 0.53) at a one-month lag, suggesting that groundwater responds to precipitation with about a one-month delay. Beyond this, the correlation weakens and becomes negative (r = - 0.41) around an eight-month lag, indicating groundwater decline due to reduced rainfall and higher temperatures. The model was validated using observations from 80 groundwater monitoring wells in Kabul city, showing a monthly correlation of r = 0.55 and a stronger annual correlation of r = 0.75. These findings confirm the reliability of GRACE-derived groundwater estimates and highlight the importance of expanding the monitoring network across Afghanistan to better capture spatial variability and support sustainable groundwater management.