At 6 rice paddy sites in 4 prefectures of Japan, we analyzed the contribution of biological N2 fixation (BNF) and ratoon rice growth to soil N fertility, combining 2-year field monitoring and simulation by the biogeochemistry model DNDC-Rice. Across the sites and years, ratoon rice was found to accumulate up to 30 kg N ha-1 without fertilization and irrigation after main rice harvest. Applying a newly built BNF model calibrated against literature data, BNF at the 6 sites were estimated to be 33-63 kg N ha-1 yr-1. Based on the simulations by DNDC-Rice under the locally typical managements, we estimated the contribution of BNF and ratoon rice to soil N fertility, being varied due to the climate, soil properties and management, as follows: (a) BNF and ratoon rice contributed 4-33% and 3-23% of N supply from soil during main rice season, respectively. (b) Contribution to main rice N uptake was 3-29% from BNF, whereas 6% or less from ratoon rice. (c) Although major part of N gain by BNF was being lost by denitrification and N leaching, BNF was contributing 88-349 kg N ha-1 (1.5-6.6%) of the organic N pool in 0-30cm soil layer. Ratoon rice was working to save N loss by reducing N leaching, consequently contributing 14-183 kg N ha-1 (0.2-3.3%) of the soil N pool. We suppose this is the first analysis that quantified the contribution of BNF and ratoon rice growth to paddy soil fertility.