The red bone marrow (RBM) exposure due to bone-seeking radionuclides can lead to grave medical consequences. In particular, the increased risk of leukemia in people exposed due to contamination of the Techa River in 1950s is associated with the RBM exposure due to ^89,90Sr. Improvement of the internal RBM dosimetry methods includes the development of computational phantoms that represent 3D models of the skeletal sites. Modeling radiation transport within such phantoms enables estimation of conversion factors from the radionuclide activity in the bone to the RBM dose rate. This paper is an extension study focused on generating a set of computational phantoms representing skeletons of individuals of different ages. The aim was to develop a computational phantom representing a 5-yearold child for internal RBM dosimetry from incorporated beta emitters. The phantoms of the skeletal sites with active hematopoiesis were created using the original Stochastic Parametric Skeletal Dosimetry (SPSD) method. With this method, every such site represented a set of smaller phantoms of simple geometric shape. RBM distribution across the skeleton, bone size, characteristics of bone micro-architecture, as well as density and chemical composition of the simulated media (RBM, bone) were determined based on the published data. As a result, a computational phantom of the major skeletal sites with active hematopoiesis representing a 5-year-old child was generated that included 43 phantoms of bone fragments. Linear dimensions of phantoms were within 3–75 mm. Micro-architecture parameters varied greatly: BV/TV ratio —13–52%, Tb. Th. — 0.09–0.29 mm, Tb. Sp. —0.48–0.98 mm.