[Objective] The aim of this study is to explore the response of soil active organic carbon loss to aggregate breakdown and transport, estimate the loss amount of active organic carbon by establishing empirical equations, and provide theoretical support for evaluating the quantitative relationship between soil organic carbon loss and carbon mineralization under water erosion. [Methods] This study took typical Lou soil in the Loess Plateau to conduct artificial simulated rainfall experiments with three rainfall intensities (60 mm/h, 90 mm/h, and 120 mm/h) and three slopes (5°, 10°, and 15°), and the loss amount of active organic carbon was estimated by establishing empirical equations. [Results] Slope gradient had the more important effect on light fraction of soil organic carbon (LFoc) loss than rainfall intensity. Both LFoc concentrations of sheet erosion and splash erosion sediments decreased first and then increased with the increase of slope gradient. Furthermore, rainfall intensity had no significant effect on the LFoc concentration of sheet erosion sediment (p＜0.05). The LFoc concentration of splash erosion sediments was obviously lower than that of sheet erosion sediments. Meanwhile, the LFocs were obviously enriched in sheet erosion sediment while those in splash erosion sediments were not obviously enriched. Compared the LFoc concentrations of aggregates with different particle sizes, the results revealed that the LFoc of ＜0.05 mm clay and silt particles and 0.05～0.25 mm aggregates were more likely to be enriched, while the LFoc of 0.25～2 mm aggregates was enriched only under light rainfall intensities and gentle slopes. As the transport of ＜0.02 mm aggregates had great correlation with LFoc loss, such a loss was well estimated by the transport amount of ＜0.02 mm aggregates (R²=0.727). [Conclusion] Runoff preferentially transports active organic carbon during water erosion, which is the reason for the large amount of mineralization of organic carbon. The loss and mineralization of active organic carbon can be well estimated by the amount of the transported aggregates in the future. [ABSTRACT FROM AUTHOR]