Abstract: Based on the biomass data of 276 sampling trees of Pinus koraiensis, Abies nephrolepis, Picea koraiensis and Larix gmelinii, the monoelement and dualelement additive system of biomass equations for the four conifer species was developed. The model error structure (additive vs. multiplicative) of the allometric equation was evaluated using the likelihood analysis, while nonlinear seemly unrelated regression  was used to estimate the parameters in the additive system of biomass equations. The results indicated that the assumption of multiplicative error structure was strongly supported for the biomass equations of total and tree components for the four conifer species. Thus, the additive system of logtransformed biomass equations was developed. The adjusted coefficient of determination (R_a²) of the additive system of biomass equations for the four conifer species was 0.85-0.99, the mean relative error  was between -7.7% and 5.5%, and the mean absolute relative
error  was less than 30.5%. Adding total tree height in the additive systems of biomass equations could significantly improve model fitting performance and predicting precision, and the biomass equations of total, aboveground and stem were better than biomass equations of root, branch, foliage and crown. The precision of each biomass equation in the additive system varied from 77.0% to 99.7% with a mean value of 92.3% that would be suitable for predicting the biomass of the four natural conifer species.