Understanding spatiotemporal dynamics and drivers of ecosystem service value (ESV) is critical for informing ecological restoration and sustainable land management, particularly in arid inland river basins. Analyzing the spatiotemporal dynamics of ESV in arid river basins and identifying key ecological and environmental drivers enable more precise diagnosis of ecological problems and provide a scientific basis for effective governance. This study evaluated the changes in ESV in the Aksu River Basin from 1990 to 2020 using the InVEST model, based on land use data, meteorological records, and biophysical parameters. A comprehensive assessment of seven key ecosystem services-including food production, water conservation, and biodiversity protection-was conducted. SHAP (SHapley Additive exPlanations) values were applied to interpret the contribution of ecological and environmental variables to ESV changes. The results showed that total ESV increased from CNY 18.904 billion in 1990 to a peak of CNY 22.323 billion in 2010, followed by a slight decline to CNY 20.806 billion in 2020. Spatially, Wensu, Xinhe, and Bachu counties exhibited substantial ESV gains, while Atushi, Akto, and Awat counties experienced significant losses. SHAP analysis identified forest quality, soil erosion, and grassland condition as the dominant drivers of ESV variation, surpassing the influence of land area alone. By combining biophysical modeling with interpretable machine learning, this study highlights the critical role of ecosystem quality rather than land area alone, offering a transferable approach for diagnosing ecological risk assessment in arid regions.