Gene therapy as a means of generating “artificial” insulin-producing cells is being considered as a potential cure for type 1 diabetes (T1D). The aim of this study was to evaluate the utility of ex vivo expanded murine mesenchymal stem cells (MSCs) as targets for gene therapy and the development of a cell replacement therapy. CD45-/Ly6+ murine MSCs were isolated from the bone marrow of non-obese diabetic (NOD) mice and nucleofected to express the bioluminescent protein Firefly luciferase (Luc2). The persistence of a subcutaneous (s.c) transplant of Luc2-expressing MSCs was assessed in immune-competent (NOD) (n=4) and immune-deficient (NOD/Scid) (n=4) animal models of diabetes. Ex vivo culture-expanded MSCs were subsequently transduced with the HMD lentiviral vector (MOI=10) to express furin-cleavable human insulin (INS-FUR), murine NeuroD1 and Pdx1; followed by characterisation of pancreatic transdifferentiation via reverse transcriptase polymerase chain reaction (RT-PCR), and acute and chronic insulin secretion assays. A s.c transplant of 1x107 and 5x107INS-FUR-expressing MSCs in NOD/Scid mice (n=5) was assessed for their ability to reverse diabetes. Luc2-expressing MSCs persisted for 2 weeks and 12 weeks respectively in NOD and NOD/Scid mice. INS-FUR-expressing MSCs lacked glucose-responsiveness and secreted human insulin chronically, whereas NeuroD1 and Pdx1-expressing MSCs lacked glucose-responsiveness and insulin secretion capabilities. Transduced MSCs did not undergo pancreatic transdifferentiation as determined by RT-PCR analysis, and upon transplantation did not reverse diabetes. The data suggests that ex vivo expanded MSCs lose their multipotent differentiation potential and may be more useful as gene therapy targets prior to expansion. This correlates with other studies where ex vivo expansion of MSCs is associated with a loss of MSC function and negative T1D clinical outcomes.