Natural ¹⁵N abundance (δ¹⁵N) in soil reflects the history of soil management. In previous studies, however, changing process of δ¹⁵N in soil, under successive applications of compost, has not been investigated. The changing process of δ¹⁵N in soil affected by successive soil management may show new aspects of δ¹⁵N in soils. To investigate the transition of δ¹⁵N values in paddy soil under successive soil management, the δ¹⁵N values in paddy soils in long-term field experiments with compost application, and the δ¹⁵N values of composts used for the experiments were measured. Soils and composts were collected from long-term field experiments with successive application of rice straw compost and from long-term field experiments with successive application of livestock manure compost. In the long-term field experiment with rice straw compost, the compost was applied at the rate of 0, 10, 20, 30, or 40 Mg ha^-1 year^-1 since 1968. For the field experiment using livestock manure compost, the compost was applied at the rate of 0 or 36 Mg ha^-1 year^-1 since 1973. The δ¹⁵N values of air-dried soils and composts were measured with ANCA-SL. The δ¹⁵N values of livestock manure composted prior to 1983 were around 6-7 permil. Values of livestock manure composted between 1984 and 1997 were approximately 11-12 permil. Values of manure composted after 1998 were about 17-18 permil. The δ¹⁵N values of rice straw compost were almost constant and averaged 5.5 permil. Variation in δ¹⁵N values in livestock manure compost reflected the degree of maturity of composting and the type of livestock manure composted. The livestock manure composts before 1983 were created by piling cattle manure on a concrete pad in the field without turning and were not well composted. Livestock manure composts from between 1984 and 1997, made by piling cattle manure in a compost house and turning it, were relatively better composted than compost from before 1983. After 1998, livestock manure composts were made from three types of livestock manure; cattle manure, swine manure and poultry manure that were stirred with a rotary mixer under forced aeration in a compost house. This composting method ensured sufficient maturity of composting. Addition of swine and poultry manure promoted ammonium volatilization which discriminated for ¹⁴N. Stability of δ¹⁵N values in rice straw compost might have resulted from the rice straw being collected from nearby and the consistent composting method, which involved piling and occasional turning in a compost house. In soils with successive application of livestock manure compost, δ¹⁵N values increased, whereas the δ¹⁵N value in soil without compost tended to decrease. The upward trend of δ¹⁵N values in soil treated with livestock manure compost was matched by the δ¹⁵N value of livestock manure compost, which became higher. Whether livestock manure compost was applied or not, soils treated with ammonium sulfate tended to show lower δ¹⁵N values than those not treated with ammonium sulfate. In soils with successive application of rice straw compost, δ¹⁵N values of the soils remained nearly the same. On the other hand, a downward trend was observed in the δ¹⁵N values of soils not treated with rice straw compost, and the trend was more pronounced when ammonium sulfate was applied. Interestingly, in paddy soil without compost and without chemical N fertilizer, the δ¹⁵N value of the soil decreased. This is attributed to the natural input of nitrogen by biological fixation and deposition with rain, which are lower δ¹⁵N values. In other words, this decrease in the δ¹⁵N value reflects the potential of paddy fields to maintain N fertility. Documentation of the transition of δ¹⁵N values of soils in long-term paddy field experiments suggests the δ¹⁵N value of paddy soil could potentially decline by natural N supply, and that the δ¹⁵N value of paddy soil is determined by the balance among the δ¹⁵N-lowering effect of natural N input, amount and δ¹⁵N value of applied nitrogen materials, and behavior of their nitrogen.