Large-scale vegetation restoration projects have significantly improved the regional ecological environment and accelerated progress towards carbon neutrality. However, the associated increase in evapotranspiration (ET) has raised concerns about water resource sustainability. Although it is widely accepted that vegetation greening leads to increased ET and associated water depletion, the hydrological impacts of extensive vegetation restoration remain incompletely understood. Focusing on the Yellow River Basin, a region of intensive vegetation restoration, we simulated two scenarios-dynamic vegetation change (DV) and no dynamic vegetation change (NO_DV)-to quantify vegetation impacts on ET and precipitation, using the modified Priestley-Taylor Jet Propulsion Laboratory model (PT-JPL), Dynamic Recycling Model (DRM) the Weather Research and Forecasting (WRF) model. The results show that the modified model, which incorporates the differences in soil evaporation and canopy transpiration under varying vegetation cover types, offered higher precision. The ET difference between the two scenarios exhibited an increasing trend of 4.51 mm per year, and the restoration of vegetation significantly suppressed soil evaporation while substantially enhancing canopy transpiration. Additionally, vegetation restoration led to an increase of 1.68 % in the annual mean precipitation recycling ratio and generated an additional 1.81 mm of precipitation per year. Mechanistic analysis revealed that vegetation restoration creates favorable hydrological conditions that enhance precipitation. These findings represent a significant contribution to understanding the hydrological response mechanisms and positive feedback loop induced by large-scale vegetation restoration and highlight that the vegetation carrying capacity in some regions may be underestimated.