In order to develop effective cell based therapies for treating several human diseases, it is necessary to increase the limited number of stem cells harvested from a single patient to be used in autologous transplantation. Along these lines, mathematical modeling may represent a useful tool for interpreting and rationalizing the in vitro expansion techniques, thus helping to find the optimal operative conditions. Specifically, this work addresses the mathematical simulation of the proliferation kinetics of sheep bone marrow mesenchymal stem cells (phenotypic characterized by flow-cytometric analysis), as model similar to the human one, seeded at different initial concentrations in petri dishes and expanded up to confluence. The sigmoidal temporal profiles of total counts obtained through classic hemocytometry are quantitatively interpreted by a novel model based on population balance approach capable to take into account the contact inhibition at confluence. First, the adjustable parameters of the proposed model are fitted against experimental data on population expansion starting from one single seeding concentration. Then, the reliability of the mathematical model is successfully tested through the prediction of cell proliferation kinetics carried out starting from different seeding concentrations. (C) 2009 Elsevier Ltd. All rights reserved.