Degeneration of the intervertebral disc (IVD) is thought to be one main factor in the development of back pain. It is not a symptom by itself but can lead to painful pathological conditions. Current treatments aim at relieving pain by conservative care, medications or surgical removal of the painful part of the disc but do not treat the causes. Disc degeneration is a progressive process which begins in the center of the disc by the loss of water due to impaired cell activity. Our hypothesis is that regeneration of the disc at the first stages of degeneration is likely to delay the progression and the need for surgery. Regeneration could be achieved by a cell-based therapy, which would bring active matrix producing and/or growth factor delivering cells. The up-scaling from bench-side to clinic for a cell therapy is limited due to safety and consistency matters. Thus, the choice of cell-type is of utmost importance to assure consistent efficacy and safety of a therapeutic application. The aim of this work was to characterize human fetal cells isolated from fetal cartilaginous tissues regarding their potential for the correction of IVD degeneration. In a first step and first chapter, a methodology for creating consistent and safe fetal cell banks from only one single tissue donation was developed. This method was illustrated by the example of fetal skin cells, which dedicated cell banks have already been tested in clinical trials. Secondly, in the second chapter, the isolation and chondrogenic potential of fetal spine cells was presented. Fetal cells were isolated from spine units (2 IVDs and 1 vertebra) from 5 donors aged from 12 to 16 weeks of gestation and amplified in monolayer to establish dedicated and consistent cell banks. Alginate bead cultures of fetal spine cells showed heterogenous matrix synthesis ability. Two fetal cell donors (14 and 15 weeks of gestation), showed good aggrecan and type II collagen production with very low type X collagen production. Mesenchymal stem cells (MSCs) can be differentiated into several phenotypes and show good regeneration capacity. Fetal cells also show good regeneration property but there is little information concerning their plasticity. Thus, in chapter three, the plasticity of fetal spine cells was investigated and compared it to that of MSCs. Cells isolated from fetal articular cartilage were also investigated as they also represent a potential source of fetal cartilaginous cells. Similarly to MSCs, fetal cells from both origins were positive for surface markers CD44, CD73, CD105 and CD166 and were negative for CD34 and CD45. Fetal cells exhibited much lower adipogenic and osteogenic differentiation levels than MSCs. As expected, fetal cells showed high chondrogenic differentiation, showed by aggrecan and type II collagen production, in TGF-β3 stimulated high cell mass density system. A mild level of type X collagen was detected in fetal spine cell pellets, whereas fetal cartilage cell pellets were highly positive. In chapter four, the response of fetal spine cells to glucose deficiency and hypoxia was assessed since these stress conditions would be similar to those of the in vivo disc. Viability of fetal cells cultured in monolayer was not altered by glucose level in normoxia and under 5% oxygen but was decreased by low glucose levels under 2.2% oxygen. Hypoxia induced an increase in SOX-9 gene expression independently of glucose level. However, a trend for a decreased gene expression of aggrecan and type I collagen under hypoxia and deficiency in glucose was observed. Finally in chapter five, fetal spine cell conditioned media (FSCCM) was tested for anti-inflammatory properties in a model of interleukin-1 (Il-1) stimulated nucleus pulposus (NP) cells. Il-1 stimulated bovine NP cells showed decreased cycloxygenase-2 (COX-2) expression and increased Prostaglandin E2 (PGE2) production in response to FSCCM, whereas human NP cells showed a decreased COX-2 gene expression as well as decreased PGE2 production. Molecular weight fractions of FSCCM had the same inhibitory effect than whole FSCCM, indicating that activity of FSCCM is more likely due to a combination of compounds rather than to a single molecule. In conclusion, fetal spine cells have been shown to be a good candidate for IVD regeneration. Fetal cell banks can easily be established and provide consistent cell lots with the extensive safety testing accomplished to assure necessary security for the patient. However, donor to donor variation in the matrix synthesis capacity will impose a strict donor selection, which could be based on spontaneous matrix synthesis in alginate bead culture. Type X collagen production was very low in unstimulated cells but could be induced by chondrogenic factors, indicating a potential risk for endochondral ossification, risk which should be assessed in an in vivo model. On the other hand, the ability of fetal spine cells to survive in low oxygen and glucose conditions, their low plasticity and their anti-inflammatory properties are interesting advantages for IVD regeneration.