Background and aimsCoal gangue-based soil reconstruction can improve the properties of damaged mine soils. However, restoring only fundamental soil properties may be insufficient for vegetation development. In addition, the soil in the semi-arid areas is scarce and relatively infertile, with targeted soil reconstruction strategies remaining undefined.MethodsThis study conducted soil reconstruction and plant development experiment, and assessed the quality of reconstructed soils. Nine soil property indicators and three vegetation indicators were selected to form the total data set (TDS). The minimum data set (MDS) was determined through principal component analysis (PCA) and correlation analysis, and two types of soil quality index (SQI) were calculated. The average correlation (ACL) was chosen to identify the comprehensive connection between a certain indicator and other indicators.ResultsCoal gangue-topsoil (CGTS) ratio and soil layer thickness influenced soil properties and vegetation development, with eight indicators showing significant differences. Soil quality achieved the level comparable to undamaged soil under 3 specific schemes. MDS included hydrolyzable nitrogen (HN), available phosphorus (AP), soil organic matter (SOM), silt content (Silt), and germination count (GC), with vegetation indicators showing high relevance and contribution. Moreover, cases of high soil quality coupled with poor vegetation development were observed.ConclusionVegetation development was not completely consistent with soil quality, and vegetation indicators provided supplementary insights for SQI. Overall, the CGTS2:1a scheme is recommended for soil quality improvement and vegetation recovery in semi-arid mining areas.