Abstract: Association classification and interspecific association analysis cultivate forest communities with specific ecological functions, such as enhanced carbon sequestration and water conservation by precise regulation of key environmental factors and species composition, thus providing a scientific paradigm for the restoration of degraded ecosystems. Associations within middle-aged pine forests in Zhejiang Province were classified using hierarchical clustering analysis. The significance and quantitative expression of interspecific associations were assessed through analyses of interspecific associations and niche overlap. Redundancy analysis (RDA) was employed to explore community ordination and the effect of environmental factors, thereby identifying the associations and the dominant environmental factors. The results revealed that the sample plots of middle-aged pine forests were classified into three associations on the basis of different dominant species. The first, Pinus taiwanensis-Pinus elliottii association, included companion species such as P. massoniana, Cunninghamia lanceolata, Schima superba, Cyclobalanopsis glauca and P. thunbergii, with shrubs dominated by Eurya japonica and Symplocos sumuntia. This association showed a competitive interspecific relationship (a positive-to-negative association ratio of 0.41), with significantly positive associations between P. taiwanensis and S. superba, C. glauca, and significantly negative associations between P. elliottii, P. massoniana and C. glauca, S. superba. Overall, the interspecific associations among the dominant species were relatively incompact. The second, Pinus massoniana-Cunninghamia lanceolata association, included companion species such as Ilex chinensis, Liquidambar formosana, Quercus fabri, S. superba, and Toxicodendron vernicifluum, with shrubs dominated by C. lanceolata. It exhibited a positive-to-negative association ratio of 0.41, with significantly positive associations between P. massoniana and C. lanceolata, and significantly negative associations between C. lanceolata and Q. fabri. The interspecific associations among the dominant species were still relatively incompact. The last, pure P. massoniana association, included companion species such as C. lanceolata, S. superba, Castanopsis sclerophylla, Q. fabri, and C. glauca, with shrubs dominated by Loropetalum chinense. It displayed a positive-to-negative association ratio of 2.6, with significantly positive associations between P. massoniana and C. lanceolata, C. glauca, and significantly negative associations between P. massoniana and L. chinense, Dalbergia hupeana. The interspecific associations among the dominant species remained relatively incompact. Nine environmental factors explained 7.9% of the species distribution variation, with elevation, mean annual precipitation, mean annual temperature and humus thickness identified as the dominant factors. The interspecific associations across associations in middle-aged pine forests followed the gradient of “competition-neutrality-weak mutualism”, reflecting the combined effects of environmental filtering and biological interactions. Regulating the interactions between dominant and companion species interactions(such as promoting the coexistence of positively associated species pairs and controlling competition intensity among negatively associated ones) and optimizing environmental factors, can support the directional cultivation of communities with specific ecological functions and the restoration of degraded pine forests, ultimately achieving multi-objective management. Future studies should focus on quantifying the functional contributions of positively associated species pairs and the thresholds for regulating negative associations to improve management precision.