Spinal cord transection results in severe neurological sequelae, and to date, there is no effective treatment. Because of the limited capacity for axonal regeneration in the spinal cord, recovery is minimal. Recently, efforts have been made to establish, by grafting neural tissue, a functional relay-station between the severed stumps of the injured cord. Previously, we used co-transplantation of neural stem cells (NSCs) and Schwann cells (SCs) to improve functional recovery of transected spinal cord. However, this effort has been partially impeded by limited neuronal differentiation of transplanted NSCs. To circumvent this problem, we have pre-differentiated NSCs toward neurons in vitro with the application of retinoic acid (RA) prior to cell grafting. Further, we genetically modified SCs to overexpress human neurotrophin-3 (hNT-3). When these cells were co-transplanted into the transected spinal cord of rats, injured animals had partial improvement (both functionally and structurally), including improved Basso, Beattie, and Bresnahan (BBB) scores, increased axonal regeneration/remyelination, and reduced neuronal loss. However, this pre-differentiation of NSCs in vitro only mildly improved neuronal differentiation of NSCs in vivo.