Abstract:[Objective] As a key area for the construction of ecological security barriers in northern China, the Mu Us Sandy Land plays a crucial role in water conservation. Understanding the temporal and spatial evolution patterns of its water conservation function is of great significance for evaluating the effectiveness of ecological restoration projects in the Mu Us Sandy Land under the context of climate change, dynamically allocating water resources, and making land use planning decisions. Previous studies have primarily focused on the relative magnitude of water conservation capacity over time and the impact of land use/cover changes on this capacity. However, research on the long-term characteristics of its water conservation function and regional contributions remains insufficient, and systematic investigations into its climate-related attribution are still lacking. [Methods] This study is based on multi-source data from the Mu Us Sandy Land during 2000–2024, including land use, soil properties, topography, and soil saturated hydraulic conductivity. Using the water yield module of InVEST model and modified formula for water conservation, it revealed the interannual dynamics and spatial differentiation characteristics of water conservation from 2000 to 2024, and, combined with the Geodetector method, analyzed the explanatory power of various influencing factors. [Results] ① The spatial distribution of water conservation in the Mu Us Sandy Land exhibited a pronounced heterogeneity, with lower values in the northwest and higher values in the southeast. ② During the study period, the water conservation depth in the Mu Us Sandy Land showed a non-significant increasing trend at a rate of 0.834mm/a, but with notable temporal and regional differences. ③ Results from single-factor analysis indicate that precipitation, soil, topography, and land use types were the most influential factors driving interannual variations and spatial differentiation in water conservation. Results from two-factor interaction analysis revealed that the combination of precipitation and land use type has the strongest explanatory power. [Conclusion] The research findings could provide a theoretical basis and decision-making reference for evaluating the effectiveness of ecological engineering in the Mu Us Sandy Land and for ecological restoration planning, and also could offer certain guidance for the scientific management of water resources and policy formulation.