Abstract: Based on the data of 160 Schima superba samples, taking DBH (D), tree height (H) and their combined variables (D²H) as independent variables of the model. Two schemes (Total Control and Algebraic Sum Control) are used to estimate the simultaneous equations of non-linear measurement errors, and weighted regression was used to eliminate heteroscedasticity. The univariate and binary independent and compatible biomass models of Schima superba were constructed. Based on the univariate compatible model, the distribution of each component biomass to the total aboveground biomass was analyzed. The results showed that the fitted results of the compatible biomass models established by the 2 schemes were good, and the fitted results of each component biomass model were basically the same. The R² value was more than 0.845, and the MPE value was less than 4.592%. Algebra and control schemes are better for comprehensive analysis. In the independent model and compatibility model, the fitting effect of binary biomass model is better than that of univariate model, and the binary model with D²H as independent variable is the best. In the 2 schemes, the distribution of biomass of each organ in the compatible biomass model is basically the same. With the increase of DBH, the ratio of biomass of trunk and trunk to total aboveground biomass increases first and then tends to be stable. The proportion of biomass of crown, bark and leaf decreases first and then tends to be stable; while DBH is more than 30 cm, the proportion of biomass of branches does not change much, the biomass ratios of trunk, trunk, bark, crown, branch and leaf were 74.5%−75.2%, 68.9%−69.2%, 5.6%−6.1%, 24.8%−25.4%, 21.5%−22.8% and 2.6%−3.2%.