Brutian pine (Pinus brutia Ten.) is the most widespread conifer species in the Eastern Mediterranean. Aboveground biomass equations for small diameter brutian pine trees are needed for accurate fuel inventory and to assess carbon sequestration potential. In this study, we developed tree biomass models based on 143 brutian pine saplings measured in 11 research plots. Aboveground biomass (AGB) was modeled with a nonlinear mixed effects model which accounted for the variability among plots. The predicted total AGB was then distributed into foliage, branch and stem components. The Beta, Dirichlet, and multinomial logistic regressions were unbiased in their estimates of biomass component proportions. The Dirichlet regression has the advantage of an additive property and does not require non-standard data.