Bone-seeking radionuclides, in particular ^89,90Sr, could get into the environment in the course of various anthropogenic radiation incidents. From there they enter a human body with food and water. This leads to red bone marrow (RBM) internal exposure. These elements were present in the composition of radioactive releases into the Techa River in 1950s, and are the major source of RBM exposure for the residents of the riverside settlements. RBM dose estimation relies on dosimetric modeling which comprises the development of 3D computational phantoms of the skeleton parts. By imitating the energy transfer in these phantoms, the conversion coefficients from the radionuclide activity in a bone to the dose rate in RBM are evaluated. The given study is yet another step in the research aimed at the elaboration of a set of computational phantoms of the skeleton for people of various age. The objective is to develop a computational phantom of a skeleton of a 10-year-old child to estimate dose to RBM due to incorporated beta-emitters. Original SPSD (stochastic parametric skeletal dosimetry) approach was used to create the phantoms. According to this method the skeleton sites containing RBM were divided into smaller segment of simple geometric shape, for which voxel phantoms were generated. The parameters for phantom generation were based on published research data. They included^ linear dimensions of bones, thickness of the cortical layer, characteristics/properties of the bone micro-architecture, density and chemical composition of the modelled media and the percentage of RBM content in bones. Generated computational phantom of the skeleton sites with active hematopoiesis of a 10-year-old child consists of 38 phantom-segments. Linear dimensions of the segments were from 3 to 88 mm, cortical layer thickness: 0.2–2.2 mm.