Abstract: The Methodology for Greenhouse Gas Voluntary Emission Reduction Project: Afforestation Carbon Sequestration (CCER-14-001-V01) points out that since the IPCC carbon measurement parameters mostly come from Europe and America, they are prone to cause uncertainties in regional carbon measurement. Therefore, local standards are preferentially adopted for standing tree biomass models. In this study, the middle-aged artificial cypress forest in Jintang County, a hilly area in central Sichuan, was selected as the research object. According to the landform, soil layer thickness and slope position, three site types were selected (thick soil layer Ⅰ in the upper slope, thick soil layer Ⅱ in the middle slope, barren soil layer Ⅲ in the lower slope), and 105 sample trees were collected to determine the carbon content and biomass of each organ, and to construct biomass model and related parameters (BCEF, WD, RSR). The results showed that: (1) The carbon content of cypress leaves was the highest and that of roots was the lowest. Among different sites, there was no significant difference in carbon content of branches, leaves and roots (P>0.05), while there were significant differences in the trunk between Ⅰ and Ⅱ, and in the bark between Ⅱ and Ⅰ as well as Ⅲ (P<0.05). The average carbon content coefficient of the whole tree was 0.4869, which was slightly higher than 0.4847 in GB/T 43648—2024. The carbon content coefficients of the aboveground part, trunk, bark of type Ⅲ and root of each site were lower than the national standard, while those of branches and leaves were higher than the national standard. (2) In the unary/binary biomass model, the optimal biomass model for each organ of cypress were mostly power functions, and a few were quadratic polynomials or exponential functions. Except for some organs of type Ⅲ, the binary model with W=a·((DBH)^b·H^c) had the best fitting effect. (3) The measured BCEF, WD and RSR were all higher than the IPCC default values. The BCEF values of types Ⅰ, Ⅱ, Ⅲ and the whole were 18.02%, 23.40%, 31.04% and 24.03% higher than the IPCC default values respectively; WD was 35.15%, 37.87%, 39.75% and 37.87% higher; RSR was 35.00%, 17.73%, 23.64% and 25.00% higher. This study could provide reference for improving the carbon measurement parameter system of the IPCC and the national standard and improving the accounting accuracy of regional forest carbon stocks.