Abstract: Earthworm casts exhibit remarkable fertility and have been used since long as an organic fertilizer for agricultural production in many areas of China. Both soil structure and properties are effectively influenced by applying earthworm casts. However, the effects of earthworm casts on the physical properties of soils from Loess Plateau have not been well understood. This study focused on the effects of earthworm cast application on soil aggregates and aggregate-associated organic carbon and their stability. Soil column experiments were conducted in the laboratory using three typical soils (Loess soil, CS; Dark loessial soil, DS; Aeolian sandy soil, AS) at four earthworm cast application rates (0, 1%, 3% and 5%). After incubation for four months, soil samples were collected from each column for determination of dry aggregate size distribution and water-stable aggregate content. Organic carbon contents of the aggregates were determined as well. The results showed that the content of dry aggregates increased after applying earthworm cast for the fractions lower than 0.25 mm by 9.2%-24.7% and 7.0%-21.3% for CS and DS, respectively. The aggregates lower than 0.25 mm in AS, however, were significantly decreased by 25.9%-34.0% under the treatments of 3% and 5% application rates. Applying earthworm cast with rates of 3% and 5% was beneficial for the increase of water stable aggregates. The contents of water stable aggregates were significantly increased as compared with CK in all three soils. The mean weight diameter, geometric mean diameter, and content of aggregates of higher than 0.25 mm increased after earthworm cast application for AS, while the value of fractal dimensions decreased. These indexes varied when the application rate reached 3% for CS and DS. The results indicated applying earthworm cast with rates of 3% helped to improve aggregate stability, while no further improvement were observed with a higher application rate of 5%. Additionally, application of earthworm casts significantly increased soil organic carbon content. Meanwhile, the contents of aggregate-associated organic carbon were higher in treatments with earthworm cast than those without, indicating a positive effect of earthworm cast application on aggregate-associated organic carbon. The relative contribution of aggregates not higher than 0.25 to the total soil organic carbon significantly decreased by 21.4%-41.1% and 15.7%-20.4% after earthworm cast application for CS and DS, but increased for AS. No significant difference, however, was obtained between the treatments of 3% and 5% application rates. This indicated applying 1% earthworm cast was able to increase soil organic carbon and aggregate-associated carbon content, but the stability of soil organic carbon wasn’t improved until the application rate reached 3%. The relative contribution of aggregates higher than 0.50 mm to total soil organic carbon was higher in CS and DS treatments, while that of aggregates in the fractions not higher than 0.25 mm and 0.25-0.50 mm to soil organic carbon was higher in the AS treatment. The contents of labile organic carbon were higher in treatments with earthworm casts than those without. As for the three soils, DS had most labile organic carbon and AS had least. Based on an integrated consideration of the stability of soil aggregate and aggregate-associated carbon, an application rate of 3% was suggested for the agricultural production. The findings of this study improve our understandings of the effects of earthworm cast application rate on soil aggregate composition and organic carbon distribution of soils in the Loess Plateau, and are able to aid in improving the application efficiency of earthworm casts as an organic fertilizer in agricultural practices in the loess area.