Mulching management has been used in many places all over the world to improve agricultural sustainability. However, the cycling of carbon in the soil under applications of mulch on sloping arable land is not yet fully understood. A four-year field experiment was carried out in Xiaofuling watershed of Danjiangkou reservoir in China. The object was to evaluate the effects of the application of straw mulch (ST) and grass mulch (GT) on dynamic changes in soil organic carbon and its fractions. Results showed that mulch applied on the soil surface increased the contents of SOC and its active fractions in the soil. Compared to the control without cover (CK), ST and GT treatments increased the contents of SOC, LOC, DOC, POC and EOC by 14.73%, 16.5%, 22.5%, 41.5% and 21%, respectively, in the 0-40 cm soil layer, and by 17%, 14%, 19%, and 30%, respectively, in the 0-100 cm soil layer. The contents of organic carbon and its active fractions decreased with increasing soil depth in all of the treatments. SOC was accumulated in the period of December to the following March. The contents of soil DOC and LOC were high in January to March, while the contents of soil POC and EOC were high in June to September. The relative contents of soil organic carbon fractions were POC > EOC > LOC > DOC over the four years. Straw mulching had no significant effect on the changes in soil organic carbon active fractions during the different periods. Based on this long-term field experiment in Danjiangkou reservoir, we found that straw mulching had a significant effect on soil, increasing SOC content and stock in slopping arable land, and that live grass mulching was more effective than rice straw mulching. We discuss possible optimal periods for the implementation of mulching practices on sloping land.