Recently, soil carbon sequestration in agro-ecosystems has been attracting significant interest as soil organic carbon (SOC) can potentially offset some atmospheric carbon dioxide. The objectives of this study were to use the RothC model to simulate soil carbon sequestration and determine the proportion of pasture production as carbon input for SOC sequestration under different pasture types and pasture management in a long term experiment established in 1992. There were two types of pastures, annual and perennial pastures, with or without application of limestone. Simulation results showed that with an initial setting for the stubble retention factor of 0.65 and root/shoot ratio of 0.5 for annual pasture and 1.0 for perennial pasture, RothC can adequately simulate SOC for both pasture types, especially annual pasture. Using an inverse modelling technique, the root/shoot ratio was determined as 0.49 and 0.57 for annual pasture and 0.72 and 0.76 for perennial pasture with and without limestone application, respectively. There was a large improvement in model performance for perennial pasture with and without limestone application. The root mean squared errors (RMSE) reduced from 3.19 and 2.99 t C ha'1 in the initial settings to 2.09 and 2.10 t C ha'1, while performance efficiency (PE) increased from 0.89 and 0.91 to the same value of 0.95 when the root/shoot ratio of 0.72 and 0.76 were used for limed and unlimed perennial pastures. However, there was little improvement for annual pasture as RMSE had little change and PE was the same. As the stubble retention factor and root/shoot ratio can be combined into one factor that measures an equivalent amount of total above-ground pasture production allocated for soil carbon input, the modelled results can be summarised as 1.2 times and 1.4 times the above-ground dry matter for annual and for perennial pasture,respectively, regardless of liming. Our results provide useful information for simulation of soil carbon sequestration under continuous pasture systems.