Abstract: It has become a trend to supplement national forest resources monitoring with forest biomass estimation. The below-ground biomass is an important component of forest biomass. Based on the below-ground biomass data of larch(Larix spp.) in northeastern China and that of Masson pine(Pinus massoniana) in southern China, two approaches commonly used for modeling below-ground biomass, i.e., below-ground biomass and root-shoot ratio equations, were compared, and problems in modeling below-ground biomass were addressed. Then, an approach to establish compatible above- and below-ground biomass equations and root-shoot ratio equation was proposed using the error-in-variable modeling method. The results show that: 1) the new approach can be used to solve the problem of “one dataset, two results”, and the estimation results using different equations are consistent and stable; 2) in case of no enough sample trees to be selected for measurement of below-ground biomass, a root-shoot ratio equation should be developed to estimate below-ground biomass with the aid of an above-ground biomass equation established through a large sample size; 3) the root-shoot ratio of an individual tree depends mainly on the diameter of the tree, and increases gradually with the growing diameter. The prediction precisions of single-tree below-biomass equations developed in this study are 85% for larch in northeastern China and 90% for the Masson pine in southern China, and can be applied for estimating forest biomass at a large scale.