Recent literature suggests that bone marrow stromal cells (BMSCs) may be differentiated into neuron-like and/or glia-like cells, implying that differentiated BMSCs may have potential use in cell replacement therapy for central nervous system disorders. However, many questions remain concerning the nature of BMSCs differentiated to express CNS antigens. For example, how long after differentiation do cells express CNS markers, and do differentiation procedures alter cell viability? This study compared neuronal differentiation methods in sister cell preparations, evaluating cell death and maintenance of the CNS antigen positivity after the Deng or Woodbury methods. Rat BMSCs were harvested, passaged, differentiated, placed in growth or maintenance media, and processed for cell viability or immunocytochemistry for NeuN and GFAP post-differentiation. We report that the Woodbury differentiation protocol produced maximally 51% neuron-like cells, yet also produced significant cell death. The Deng differentiation method produced 13% neuron-like cells and without marked cell death. No significant increases in GFAP immunoreactivity (IR) were seen after differentiation by either protocol. Following both protocols, removal of cells from the maintenance media significantly decreased expression of NeuN. Thus, differentiation procedures may be substantially affected BMSC potential, and maintenance of immunoreactivity to neuronal antigens was dependent on specific, nonphysiological environmental conditions.