Underground mining (as opposed to open‐cast) often causes large‐scale subsidence, leading to various types of disturbances to surface vegetation. Adequate quantitative assessment of the long‐term effects of underground mining on the growth of different plant communities is important and still lacking. To address these issues, a vegetation growth contract model (VGCM) was proposed, and six indicators including the growth trend (GT), annual growth (AG), normalized spectrum entropy (Hsn), as well as the average value of annual‐average normalized difference vegetation index (NDVI; ANDVIave), annual‐maximum NDVI (ANDVImax), and annual‐minimum NDVI (ANDVImin) were selected. The long‐term effects of underground mining (EM) on the herb, shrub, and tree communities in the Nanjiao mining area, China, from 1987 to 2017 were evaluated. The results show that the plant communities, which maintained the same type in the areas influenced and not influenced by mining, accounting for 48.07% and 46% of the total area, respectively. As for these plant communities, underground mining had a significant negative effect on the AG, ANDVIave, and ANDVImax of both the herb and tree communities, while it had a positive effect on the GT and Hsn of the shrub community. Overall, underground mining had a negative effect on these three types of plant communities, and the EMs of the herb, tree, and shrub communities were −15.10, −6.79, and −4.03%, respectively. This research could provide a reference for evaluating the long‐term effects of mining activities on vegetation, and also give more insights into the effects of underground mining on different plant communities.