Study region: Mbakaou and Bamendjing basins (Sanaga River sub-basins). Study focus: In this study, the availability of water resources was assessed over the period 2002–2019, based on the SWAT (Soil and Water Assessment Tool) hydrological model and certain meteorological and spatial reference data available for the region (Merra2, Landsat, etc.). Forecasts of its evolution were then made with the same tool (SWAT) over two futures periods (near 2024–2035 and medium: 2036–205) based on data from four (04) regional climate models (RCMs) (CCCma, HIRHAM5, RCA4 and REMO) and future land use and land cover (LULC) data simulated using the CA-Markov procedure. To separate the impact of climate variability (CV) and land use and use and land cover changes (LULCCs) on future water resources, two evolution scenarios (experiments) were established: (1) the impact of the CV, by associating future climate data with LULC from the historical period; (2) the impact of LULCCs, by combining future LULC maps with climate data from the historical period. New hydrological insights for the region: The performances of the SWAT model are satisfactory in calibration and validation on the two basins with R2, NSE and KGE greater than 0.68. Two models (CCCma and REMO) predict a decline in water resources in these basins, and two others (HIRHAM5 and RCA4) the opposite. The REMO model seems the most reliable. It predicts a drop in precipitation and runoff (SURQ) in the two basins that do not respectively exceed –19% and –31%. CV is the only forcing whose impact will be visible in the dynamics of future water resources, given the insignificant changes expected in the evolution of LULC patterns. The results of this study could contribute to improving the management of water resources in the studied basins and the region.