Abstract: Soil stores the largest carbon pool in terrestrial ecosystems, and thus accurately assessing its size and spatial distribution pattern is a critical scientific question in soil ecology study. Previous studies have shown that soil carbon density generally decreases along the soil profile from top to bottom; however, the regulatory mechanisms controlling this attenuation process are still not well understood. Based on field measurements from 80 sampling sites in a subtropical evergreen broadleaf forest of Chebaling, Guangdong province, this study systematically evaluated the spatial variation of the 0–100 cm soil carbon pool and its decay coefficient β along the soil depth. Results show that the average soil carbon storage in the Chebaling subtropical evergreen broadleaf forest is 11.09 kg C m⁻², and the mean decay coefficient β for soil carbon density along soil profile is 0.9611. By using the least subsets square regression and structural equation modeling analyses, we found that β was dominantly determined by surface soil fine root biomass in this subtropical evergreen broadleaf forest. Specifically, greater fine root biomass in the surface soil contributed to a smaller β, indicating greater carbon accumulation in the surface soil. Other factors, including surface soil carbon-to-nitrogen ratio, total ambient radiation, and aboveground biomass, are also contributed to the variation of β. These factors totally explain 66% of the spatial variation in β. By combining the simulated β values based on these factors with surface soil carbon pool can effectively predict the 0-100 cm soil carbon pool in the subtropical evergreen broad-leaved forests. In conclusion, β and whole-profile soil carbon storage could be well predicted using plant and surface soil characteristics. This approach has great potential in extrapolating the abundant surface soil carbon data to estimate soil carbon stocks down to deep soil layer, thereby providing an efficient pathway for estimating whole-profile soil carbon storage in in subtropical evergreen broad-leaved forests lacking deep soil samplings.