While intervertebral disc (IVD) degeneration is associated with the majority of cases of low back pain, current treatments are symptomatic rather than curative. Tissue engineering offers a treatment that both cures the problem of disc degeneration and restores normal disc function. One of the major problems for any tissue engineering strategy, however, is ensuring that both the cells and matrices used are suitable for the target tissue. In this study, we have developed and studied a potential system for tissue engineering of the nucleus pulposus (NP) of the severely degenerate IVD. While cells from degenerate discs are not suitable for tissue engineering, bone-marrow-derived mesenchymal stem cells, which are capable of differentiating into chondrocyte-like cells such as those found within the NP of the disc, offer a potential source of cells. We have used transfection with adenoviral SOX-9, a transcription factor involved in differentiation of MSCs along the chondrogenic lineage, combined with culture in a specialised medium, to differentiate monolayer MSCs to NP-like (chondrocyte-like) cells, as shown by real-time quantitative polymerase chain reaction for NP-marker genes. We have also replicated these findings on porous, biodegradable three-dimensional (3D) poly-l-lactic acid scaffolds and shown expression and deposition of NP matrix markers such as type II collagen and aggrecan. We are therefore proposing pre-differentiation of human MSCs and seeding on porous, biodegradable 3D synthetic polymer scaffolds as a realistic tissue engineering strategy for regeneration of the degenerate human IVD.