Objective Soil organic carbon (SOC) mineralization is generally measured by laboratory incubation of loose soil samples. However, the structure of loose soil samples is of great difference from that of bulk soil samples. The relationship between SOC mineralization of loose soil samples and bulk soil samples is not clear. Soil samples close to field conditions can be obtained by repacking soil columns. Nevertheless, compactness can affect soil pore structure and may influence SOC mineralization. Therefore, this study aimed to evaluate whether it is accurate to represent SOC mineralization in the field by incubating loose soil samples and how compactness influences soil pore structure or SOC mineralization in repacked soil columns.Method Soil samples were collected from a long-term field experimental site with treatments receiving different amount of pig manure. In our first incubation experiment, all of these soils were selected and two treatments were set up in each soil: loose soil samples and repacked soil columns. In the second incubation experiment, only one soil was used, and the soil was repacked into columns with four bulk densities, which were 1.1(BD_1.1), 1.3(BD_1.3), 1.5(BD_1.5) and 1.7(BD_1.7) g·cm^–3. The samples of these two experiments were incubated for 57 d and 28 d, respectively. SOC mineralization was measured during incubation, and soil pore structure was quantified using X-Ray micro-computed tomography (μ CT) imaging.Result At the end of incubation(57 d), the cumulative amount of SOC mineralization was significantly different between loose soil samples and repacked soil columns. The cumulative amount of SOC mineralization in the loose soil samples was about 4 times that of the repacked soil columns. In the second experiment, the total porosity decreased by 12.9%, 14.8% and 17.4%, respectively under BD_1.3, BD_1.5 and BD_1.7 compared with BD_1.1. In relative to BD_1.1, the increase of compactness decreased macro-porosity by 19.0%, 65.5% and 88.5%, respectively under BD_1.3, BD_1.5 and BD_1.7. In addition, the water-filled pore space (WFPS) increased from 36.4% to 91.8% and air-filled pore space (AFPS) decreased from 63.6% to 8.2%. At the end of incubation(28 d), the cumulative amount of SOC mineralization generally increased as bulk density increased up to 1.5 gžcm^–3, after which there was a decrease. The regression analysis showed that there was a significant nonlinear relationship between the cumulative amount of SOC mineralization and total porosity, macro-porosity, WFPS and AFPS. The cumulative amount of SOC mineralization increased with increasing total porosity and macro-porosity until a level of 46% and 3.7% was respectively reached, afterwards it began to decline. Also, the relationship between the cumulative amount of SOC mineralization and WFPS and AFPS showed the same trend. The cumulative amount of SOC mineralization was the highest when WFPS was 66% or AFPS was 34%.Conclusion Laboratory incubation using loose soil samples will overestimate the potential of SOC mineralization in the field, while a change of compactness will modify soil pore structure and subsequently affect SOC mineralization. There is a significant nonlinear relationship between the cumulative amount of SOC mineralization and porosity.