Vegetation is a core component of terrestrial ecosystems, thus understanding the drivers of its dynamics is crucial for ecological conservation and management, especially in regions rich in natural resources but ecologically fragile. Here, we examined vegetation cover changes in China's western frontiers and quantified the relative importance of key drivers of vegetation dynamics. First, we employed the Dimidiate Pixel Model (DPM) via Google Earth Engine (GEE) to estimate fractional vegetation cover (FVC), followed by trend analysis using the Theil-Sen median and Mann-Kendall methods to examine FVC dynamics from 2010 to 2021. Next, we applied the optimal parameter-based geographic detector (OPGD) to further assess the impact of natural (ie., temperature, precipitation, and elevation) and human (ie., land use and population density) factors on FVC. Our results revealed that approximately 61.26% of the vegetation-covered regions in China's western frontiers have shown improvement. In Xinjiang, population density, land use, and precipitation were the primary drivers of FVC changes, with Q values exceeding 0.20. In Xizang, precipitation, elevation, temperature, and land use changes were the dominant drivers, with Q values greater than 0.30. Furthermore, interactions between natural and human factors significantly influenced FVC variation. Our findings have the potential to provide references for promoting sustainable vegetation management in western China.