【Objective】Nitrogen (N) deposition is an important global climate change factor affecting soil phosphorus (P) cycling in forest ecosystems. However, understory N application typically applied directly to the ground, has been widely used to simulate N deposition in forest ecosystems in the past. This simulation method may neglect the retention and adsorption effect of the forest canopy, which cannot truly assess the effect of N deposition on soil P cycling. Moreover, organic N is another important component of atmospheric N deposition, but the ecological effect of the organic N deposition has not been fully studied.【Method】To evaluate the effect of atmospheric N deposition on soil P cycling, we conducted a field experiment involving six treatments: canopy control (CNA-CK), canopy inorganic N addition (CNA-IN), canopy organic N addition (CNA-ON), understory control (UNA-CK), understory inorganic N addition (UNA-IN), and understory organic N addition (UNA-ON) based on the Anji Moso bamboo ecosystem research station of Zhejiang A&F University, in Anji County, Zhejiang Province. The N deposition rate was set at 50 kg·hm-2·a-1 (based on the atomic mass of N, the same as below). IN uses ammonium nitrate while ON uses a mixture of 25 kg·hm-2·a-1 urea and 25 kg·hm-2·a-1 glycine as the N source. The CK treatment involves adding an equal amount of water. Soil P fractions, microbial biomass P, acid phosphatase activity, P cycling functional gene abundance, and some physicochemical properties were measured to investigate the effects of different N deposition simulation approaches and N components on soil P fractions and their driving factors.【Result】The results showed that canopy N addition (CNA) significantly reduced soil total P, occluded P, and labile P concentrations compared to understory N addition (UNA) with the percentage of 15.1%-26.5%, 18.3%-21.5% and 9.7%-38.3%, respectively. However, soil P fractions did not differ significantly between CNA-ON and CNA-IN treatments, whereas UNA-IN treatment significantly reduced resin P and labile P content compared to UNA-ON treatment. Acid phosphatase activity and pH were the main factors affecting soil P fractions, but N deposition did not significantly influence soil P cycling functional gene abundance.【Conclusion】We, therefore, suggest that simulating N deposition via CNA significantly reduced the contents of total P and P fractions such as occluded P and labile P of Moso bamboo forest soils, whereas simulating N deposition via conventional UNA underestimated this reduction effect. Soil P is closely linked to the cycling of soil carbon and N, which are vital for maintaining a balanced nutrient ecosystem. Consequently, future simulation experiments on N deposition should systematically consider the effects of N deposition simulation approaches and N addition components on soil carbon, N, and P cycling processes.