Paddy soil is a key type of cultivated soil in China. Precisely understanding the characteristics and main control factors of paddy soil organic carbon fractions is critical to quantitatively evaluating soil organic carbon (SOC) quality and monitoring its trends. We collected 65 soil samples from 13 soil profiles associated with three major paddy soil types in the Yangtze River Delta, including 5 hydromorphic paddy soil profiles, 3 percogenic paddy soil profiles, and 5 degleyed paddy soil profiles. Through long-term soil incubation experiments, the SOC decomposition amounts were measured at different times (1, 3, 5, 7, 10, 15, 22, 29, 36, 43, 50, 60, 70, 84, 99 d); the amount of resistant SOC(C_r) was determined by the acid hydrolysis method, and the amount of active SOC(C_a), slow SOC(C_s), and resistant SOC(C_r) were simulated by fitting a three-pool first-order equation to the above data. Distribution characteristics of paddy soil organic fractions in profiles (0-100 cm) were analyzed and illuminated, and main control factors on SOC fractions were obtained through principal component analysis. Finally, a regression model was established to predict SOC fractions from the main control factors. Results showed that the amount of active SOC(C_a), slow SOC(C_s), and resistant SOC(C_r) declined with the increase of soil profile depth, and that the rate of SOC fraction decrease in the upper layer (0-40 cm) was faster than in the subsoil (40-100 cm). The type of paddy soil did not influence the amount of soil organic carbon fractions significantly. The C_a pool comprised less than 5.3% of the total SOC, and the proportion of the C_r pool, which was more than 60% in total SOC, was larger than the combined proportions of the C_a and C_s pools. The amount of SOC fractions was significantly higher in topsoil (0-20 cm) than in other soil horizons. The variation in total SOC was mainly due to the C_s and C_r contributions. Therefore, more attention should be paid to the fractions of C_s and C_r when maximizing carbon sequestration in soils. The type of paddy soil and the depth primarily influenced the organic carbon composition of the topsoil. This research found that total SOC, total nitrogen (TN), and pH were the main control factors influencing the differences in SOC fraction amounts, and they can be used to predict SOC fraction amounts to more comprehensively understand the SOC cycle. Determining the amounts and composition proportions of SOC fractions can contribute significantly to mastering soil organic carbon pool dynamics. Creating a cost effective model to predict SOC fractions is meaningful and urgent. According to our research, SOC fractions can be predicted from the basic physical and chemical properties of soils.