Elucidating the variation in leaf functional traits and the trade-off relationships among those traits are of great significances to reveal the response mechanism's and adaptation strategies of plants towards changing environment. The variation of plant functional traits organically links the species adaptation strategy with the ecosystem functions and processes, and the ranges including interspecific and intraspecific variations vary with species and are closely related to environmental gradients. Meanwhile, the leaf economics spectrum (LES) at the global scale can reflects the trade-off relationships between the main structural, chemical and physiological traits. However, local community puts environmental factors and functional traits in the same system, and the main reasons responsible for the variation in functional traits and the applicability of LES still need to be deeply investigated. In this study, we collected 8 leaf functional traits of 6 dominant tree species in a 1 ha permanent monitoring plot of a Lithocarpus glaber-Cyclobalanopsis glauca subtropical evergreen broad-leaved forest, and the studied traits included leaf area (LA), specific leaf area (SLA), leaf dry matter content (LDMC), leaf thickness (LT), leaf carbon (LC), leaf nitrogen (LN), leaf phosphorus (LP) content and carbon-nitrogen ratio (LC :LN). We used multiple comparison and principal component analysis (PCA) to explore the variation in leaf functional traits at the interspecific, intraspecific and life form levels, and to analyze the relationship among those functional traits. The results showed that:(1) Leaf functional traits were significantly different at the intraspecific and interspecific levels in this community. There was a moderate variation with a coefficient variation (CV) ranging from 0.02 to 0.59, and the structural traits were more conservative than chemical traits. The variation pattern of leaf functional traits conformed to "a spatial trait variance partitioning hypothesis". Leaf functional traits significantly differed among life forms, with lower SLA and LA for coniferous tree species compared to those for broad-leaved tree species, lowest LC and LDMC for evergreen trees, yet the deciduous tree species exhibited the highest SLA, LN and LP along with the lowest LT and LC :LN. (2) Life form and interspecific differences contributed the most to the variation in leaf functional traits, and intraspecific differences also contribute significantly to those variations with lower contribution rate. Life forms had relative higher contribution to the variation in most functional traits with the contribution rate reaching 93.11% and 91.76% for LDMC and LC:LN, respectively. The contribution rate of interspecific variation (except LDMC) to structural traits was higher than that to chemical traits. Intraspecific variation had a higher contribution rate to LP (23.66%) than interspecific variation. (3) There was a significant correlation between leaf functional traits. Conifers and broad-leaved trees clustered in different areas of PCA ordination chart. The leaf economics spectrum was applicable in the subtropical evergreen broad-leaved forest, and diverse tree species could coexist in this community by the trade-off strategies among leaf structural and chemical traits. The results could provide a scientific basis for understanding the adaptation strategies of diverse tree species to environment, predicting the dynamic changes of community and formulating vegetation restoration strategy.