The definition of stem diameter limits between pulp production and residues recovered from harvesting for energy can influence the dry biomass content of forest products. In this context, the aim of this study was to build linear regression models employing categorical variables to accurately estimate the biomass of pulp and residues from Eucalyptus saligna and E. urophylla × E. grandis clonal stands. Forest inventory and dry biomass data from thirty trees distributed in diameter classes of each stand were used as a database. The measurements of pulp dry mass and residues were treated as dependent variables. Concurrently, variables such as diameter at 1.3 m above the ground, total height, and transformations and combinations of these variables were considered as independent variables. The stem diameter limits of 8, 10, 12, and 14 cm along the stems were considered as categorical variables. To predict pulp and energy biomasses based on stem diameter limits, the categorical variable assumes a value of 1 (indicating an estimate within a particular k-factor), while the other k-factors were set to 0 (zero). The fitted models allowed for the estimation of dry biomass within the forest stands, resulting in percentage errors of estimation per hectare ranging between 4.5% and 11.2%. Moreover, these models incorporating categorical variables facilitated the simplification of estimates for both pulp biomass and recovered residues within commercial Eucalyptus stands.