AbstractUsing process‐based models to predict changes in carbon (C) stocks enhances our knowledge on the long‐term dynamics of soil organic carbon (SOC) in various land management systems. The objective of this study was to apply the Century model to evaluate temporal SOC dynamics in two temperate intercrop systems [1:2 (one row of maize and two rows of soybeans); 2:3 intercrop (two rows of maize and three rows of soybean)] and in a maize and soybean sole crop. Upon initiation of intercropping, SOC increased by 47% after ≈ 100 years, whereas SOC in the maize sole crop increased by 21% and 2% in the soybean sole crop. The quantity of crop residue input was sufficient to increase the active (turnover time of months to years) SOC fraction in the intercrops and the maize sole crop, but not in the soybean sole crop. The slow fraction, with a turnover time of 20 to 50 years, increased in all crop systems and was the major driver of SOC accumulation. A 3 to 15% loss of SOC from the passive fraction, with a turnover time of 400 to 2000 years, in all crop systems showed the long‐term impact of land‐use conversion from historically undisturbed native grasslands to intensive agricultural production systems. This study provided an example of the potential of process‐based models like Century to illustrate possible effects of cereal–legume intercropping on SOC dynamics and that the model was able to predict SOC stocks within –7 to +4% of measured values. We conclude, however that further fine‐tuning of the model for application to cereal–legume intercrop systems is required in order to strengthen the relationship between measured and simulated values.