Abstract: Machilus pauhoi is an excellent local hardwood species with ecological value and a variety of economic functions in subtropical forest ecosystem. In order to estimate the biomass simply, conveniently, and accurately, the best leaf， stem, root and total biomass estimation models based on canopy diameter (C), diameter at ground level (D) and plant height (H) were chosen by using regression analysis and model selection method. The results showed that the correlation coefficient between biomass and CH (canopy diameter multiply plant height), C²H variable was the biggest and W=b₀x^b₁ was the best model for estimating biomass. Compared with the models with D²H (basal diameter square multiply plant height), the power curve models with C²H as variable had the same estimation precision. With the convenience in real forest investigation, the models with C²H as variable were better than those with D²H as variable. The results should be further validated to find out whether they fit trees from different plant densities or the estimation of the large individual biomass or not.