Abstract: For theoretical reference on low carbon, high profit and efficient agriculture, a long-term organic fertilizer experiment was conducted since 1981 to study the effects and economic benefits of different organic fertilizers, fertilizer doses and application methods on carbon emission and carbon sink in paddy field ecosystems. Treatments of non-fertilizer (control), Astragalus sinicus application in early rice (15 t·hm^-2) (M1), double amount of A. sinicus application in early rice (30 t·hm^-2) (M2), A. sinicus application (15 t·hm^-2) plus pig manure application (15 t·hm^-2) in early rice (M3), A. sinicus application in early rice (15 t·hm^-2) plus pig manure application in late rice (15 t·hm^-2) with straw mulching (4 500 kg·hm^-2) in winter (M4), and NPK-chemical fertilizer in both early and late rice (NPK) were set up in the experiment. The soil samples were collected once every five years to measure organic carbon content after late rice harvest. Then rice biomass and yield were measured once every five years to evaluate the economic and carbon costs/benefits (5-year average) of the ecosystem after early rice and late rice harvest. Results showed that compared with the control, M1, M2, M3, M4 and NPK treatments significantly increased rice yield (P<0.05) in a range of 30.88%-96.52%. Increase in the years promoted rice yield most under M4 treatment. Long-term organic fertilizer application significantly increased SOC (soil organic carbon) content and soil carbon sink ability. Soil carbon sink of M2, M3 and M4 treatments were significantly higher than that of M1, NPK and CK treatments. Crop carbon sink under long-term organic fertilization treatments, which was 6.76-8.83 t(C)·hm^-2·a^-1 for double-cropped rice, was improved greatly. Compared with the control, net carbon sink under M1, M2, M3, M4 and NPK treatments increased significantly (P<0.05) with increment of 1.43-3.93 t(C)·hm^-2·a^-1. Carbon emission caused by production activity of each treatment remained unchanged for different years of fertilizer application. The differences in net carbon sink among treatments were mainly caused by variation in carbon sink of ecosystem, whose changing trend was similar to that of rice yield. Long-term organic fertilizer application significantly reduced chemical fertilizer input, but also significantly increased the economic benefits of double-cropped rice (P<0.05) to a maximum of 25 683.7 ￥·hm^-2·a^-1 (under M4 treatment). In conclusion, long-term organic fertilizer application significantly increased soil carbon sink and economic benefits. Besides, an integrated application of A. sinicus, pig manure and crop straw was obviously advantageous over sole application of A. sinicus in terms of increasing net carbon sink effects and economic benefits of paddy field ecosystem.