Abstract: The illegal mining activities in the Muli Coalfield have severely damaged the ecological environment of the Qilian Mountains, which has aroused great concern of the Party Central Committee. This study analyzes the spatiotemporal differentiation patterns of vegetation coverage in the entire Muli Coalfield and its four sub-mining areas (Jiangcang, Duosuo Gongma, Juhugeng, and Hushan) from 2019 to 2023, using Normalized Difference Vegetation Index (NDVI) remote sensing data. By integrating spatial-temporal comparison analysis, zoning statistics, trend analysis, and change detection methods, it evaluates the vegetation restoration outcomes of ecological restoration projects. Through regression analysis, residual analysis, quantile-based threshold classification, and comaparison analysis, the study dissects the driving mechanisms of NDVI variations spatially and temporally. It quantitatively identifies regions where human activities exert significant influences on NDVI and clarifies the net effects and regional disparities of ecological restoration interventions. The results show that: The average NDVI across the entire region increased from 0.58 to 0.62, with an annual growth rate of 6.9%. Among all sub-regions, the Juhugeng mining area showed the most significant vegetation restoration effect: its average NDVI rose from 0.38 to 0.51, and the proportion of area with high vegetation coverage (NDVI > 0.5) increased from 36.7% to 54.2%. The area with significant NDVI improvement between consecutive years exhibited a trend of initially increasing and then decreasing: the proportion of areas with significant NDVI increase in the Juhugeng and Jiangcang mining areas peaked during 2020—2021, reaching 26.1% and 13.1% respectively; while the Duosuo Gongma and Hushan mining areas reached their maximum values during 2021—2022, with proportions of 21.7% and 10.7% respectively. In terms of impact mechanisms, topography had stronger explanatory power for the spatial differentiation of NDVI than climatic factors. The proportions of areas where human activities exerted positive impacts on the temporal changes of NDVI in Juhugeng and Duosuo Gongma mining areas were 25.1% and 11.6% respectively, significantly higher than that in the surrounding areas (5.0%), highlighting the dominant role of ecological restoration projects in improving vegetation coverage. The proportions of significant negative impacts in the above-mentioned mining areas (4.4%, 6.2%, 4.5%) were close to that in the surrounding areas (5.0%), indicating that ecological restoration projects did not additionally exacerbate vegetation degradation. In summary, the ecological restoration projects in Muli Coalfield significantly improved vegetation coverage in the mining areas, but there existed regional heterogeneity and ecological vulnerability. The negative impacts of human activities were within a controllable range and did not pose a significant obstacle to the overall process of ecological restoration. The ecological restoration of Muli Coalfield is not a one-time effort. In the future, it is necessary to formulate differentiated dynamic monitoring and maintenance strategies for different mining areas, with special attention to areas with low vegetation coverage and severe degradation. It is essential to explore the causes, adjust restoration measures in a timely manner, and strictly control the negative interference of other human activities to effectively cope with potential risks of ecological degradation and continuously consolidate the achievements of ecological restoration. These research outcomes not only provide practical technical methodologies for monitoring vegetation changes and evaluating restoration effectiveness in mining areas but also offer a replicable technical paradigm for ecological restoration efforts in the Muli Coalfield and other alpine mining regions.