Increased nitrogen (N) deposition due to anthropogenic activities has been recognized to have important effects on forest carbon dynamic and plant diversity. However, most N manipulative experiments from the understory have suggested that external N addition can influence soil organic carbon pool, very little is known about the effects of canopy N addition on soil organic carbon fractions. To investigate these effects, a field experiment was carried out with four different treatments including control (CK), canopy N addition (CN, 25 kg N hm^-2a^-1, understory removal (UR), and canopy N addition plus understory removal (CNUR, 25 kg N hm^-2a^-1) in a Chinese fir plantation at Shaxian State Forest Farm, Fujian Province. The experiment was started in June 2014 and was a random block design with four replicates. After 5-years treatment, soil samples were collected by a stainlessness core (3.5 cm diameter, 60 cm length) and separated into three depths of 0-10 cm, 10-20 cm, and 20-40 cm. Then, total organic carbon (TOC), recalcitrant organic carbon (ROC), readily oxidizable organic carbon (ROOC), particular organic carbon (POC), microbial biomass carbon (MBC), and water-soluble organic carbon (WSOC) were determined. The results showed that soil TOC and ROC contents in the CN treatment had no significant changes, but the contents of soil ROOC at the 10-20 cm layer and MBC at the 20-40 cm layer, and the proportion of WSOC to TOC in surface soil layer significantly decreased and increased, respectively, compared to CK treatment. Moreover, single UR treatment did not influence soil organic carbon components, while CNUR treatment significantly decreased soil WSOC contents and proportions to TOC. Pearson analysis showed that both soil labile and recalcitrant organic carbons were positively correlated with soil moisture, water-soluble organic nitrogen, microbial biomass nitrogen, and ammonium nitrogen. These results indicated that soil labile organic carbon (LOC) was more sensitive to the short-term CN (5 years) compared to ROC, and the responses of subsoil (10-40 cm) were larger than top soil (0-10 cm). CN could promote soil LOC decomposition and consequently have a negative effect on the accumulation of soil organic carbon. However, UR was likely to offset the decrease of soil LOC induced by CN in the short time since UR would change the soil water and available nitrogen to decline the decomposition and transformation of soil organic carbon. All these findings suggest that the effects of CN and UR on soil organic carbon fractions and dynamics are complex, which requires a much longer-term manipulative experiment to understand the responding mechanism of soil carbon dynamics in the future.