Abstract: Background, aim, and scope In the context of climate change, the warm and dry climate continues to spread across the subtropical regions. To understand the wet and dry evolution of the past climate, different moisture indicators need to be investigated. Research on relative humidity (RH) helps to better understand the response of trees to moisture signals in the subtropics. In this study, a tree-ring width chronology was applied to reconstruct the RH of the Dabie Mountains (DBS) from previous April to previous July. Materials and methods The tree-ring width chronology of Pinus taiwanensis in Tiantangzhai (31.11°—31.12°N, 115.7°—115.8°E, 1550 m a.s.l), DBS, was obtained from 47 cross-dated tree-ring width measurements (Cai et al, 2018). Pearson correlation analysis was adopted to analyze the relationship between tree-ring width growth and various climate signals, e.g., RH, SPEI (Standardized Precipitation-Evapotranspiration Index), PDSI (Palmer Drought Severity Index) and soil humidity. RH was selected for reconstruction by using a simple linear regression equation. The high-frequency relationship between the reconstruction and the observed data was subjected to the first-order difference test. The reliability of the reconstruction was tested using the Bootstrap and Jackknife methods. Spatial analysis (http://climexp.knmi.nl/) and comparative analysis were performed to explore the spatial and temporal representativeness of the reconstruction. Results Current study revealed that the negative correlation between tree-ring width index and RH from previous April to previous July (RH4—7) was statistically significant (r = -0.68, p<0.01). Based on the simple linear regression equation (RH4—7 = -7.4106×STD + 83.2145), April—July RH for the period 1846—2010 was reconstructed. The reconstruction explained 46.4% of the instrumental variance (45.3% after adjusting the degree of freedom). Three wet periods (1891—1903, 1905—1921, and 1950—1993) and three dry periods (1868—1890, 1922—1934, and 1994—2005) were identified from the reconstruction. Discussion At high altitudes, the negative correlation between RH and tree-ring width might be driven by sustained high temperatures in summer which increased the relative humidity of the air. Since higher water vapor pressure suppressed evapotranspiration, the sensitivity of trees to moisture was reduced. Comparisons of the reconstruction with April—June RH in the Tianmu Mountains (TMS), Zhejiang Province and the dry/wet index (DWI) near the study area (Wuhan, Hefei, and Anqing) indicated high coherence at decadal scales. The correlation coefficient between our reconstruction and TMS April—June RH was 0.32 (n = 165, p <0.01). On the 11-yr (r = 0.68, p <0.01) and 21-yr scales (r = 0.74, p <0.01), the wet/dry variations of the two places were relatively consistent. Also, at 11-yr and 21-yr scales, the reconstructed RH significantly correlated with the DWI record, with r value of 0.42 and 0.42, respectively (p <0.01). Conclusions The positive correlation between RH and tree-ring width chronology in the DBS from previous April to previous July was statistically significant. The reconstructed RH indicated evident annual to decadal variations during 1846—2010. At decadal scale, the reconstructed RH4—7 in the DBS could represent the moisture condition of the southeast China over the period 1846—2010. Recommendations and perspectives RH of the DBS reconstructed using the tree-ring width index should provide a reference about the history of dry and wet evolution in the southeast China, and confirm its subtropical aridity trend in recent decades.