CO₂ flux from the soil to the atmosphere represents a major flux in the global carbon cycle. But there remains some limitations on predicting this flux because of multiple control mechanisms that interact at temporal and spatial scales. Soil can be described in successive layers and the processes and soil properties were described separately for each layers. The CO₂ flux between the layers was driven by diffusion, which depended on CO₂ concentration, porosity, soil temperature and moisture content. Soil CO₂ flux and its major driving factors, especially in vertical profile, may change with great temporal and spatial variability. Compared to deep soil, soil surface have great difference in temperature and moisture content. Therefore, investigation of the differences and influencing factors of soil CO₂ flux changed with soil depth could be essential for accurate evaluation of soil carbon emission. At present, the soil CO₂ vertical gradient measurement method is rapidly gaining popularity because it allows to continuously and automatically measure soil CO₂ flux at different temporal scales with minimal disturbance to the natural soil structure. It is also comparable with other methods. The soil CO₂ gradient method uses Fick's first law to calculate soil CO₂ efflux, relying on both measurements of soil CO₂ profile and on the CO₂ diffusion coefficient in the soil (Ds). Determining the latter with confidence is a challenge. Modelling as an approach can potentially be used to determine Ds. It is possible if the detailed information on some important soil properties (i.e. total soil porosity and air-filled porosity) are available. In this study, Fick's diffusion method and five kinds of diffusion coefficient models, including Penman, Marshall, Quirk, and Moldrup, were used to estimate soil CO₂ flux at different depth of the Chinese fir Plantation (Cunninghamia lanceolata) in a Base of Forest Management and Technology in Sanming (26°191, 117°361), Fujian Province, China. The modeling soil CO₂ flux were also compared with the results measured by Li-8100 automated soil CO₂ flux system. The results were as follows:(1) Calculation of surface (at 5cm) CO₂ flux estimated by 5 kinds of diffusion models had significant correlation with Li-8100 measurement result (P<0.01). Among the models obtaining diffusion coefficient, soil CO₂ efflux measured by Moldrup model were the closest to the results of the soil chamber; (2) CO₂ flux in soil different depths had obvious daily variation patterns, and the mean flux were (2.17±0.06),(2.18±0.13)、0.54±0.04、(1.65±0.20) and (1.74±0.04) μmol m^-2 s^-1 at 5, 10, 20, 40, and 60cm depth respectively (mean±SD); And (3) by plotting CO₂ efflux vs. soil temperature, we found that CO₂ efflux correlated exponentially with soil temperature at the depth of 5cm, 10cm, 60cm, with Q₁₀ of 1.35, 2.01, 4.95, respectively. But soil water content was not significant correlation with soil CO₂ efflux. Using CO₂ solid concentration detector can automatically and continuously monitor CO₂ concentration in soil of different depth soil. The estimated CO₂ efflux using this method was close to chamber measurements, suggesting that this method can be use for long-term continuous measurements of soil CO₂ efflux.