Regenerative therapies are promising treatments for early intervertebral disc degeneration. To test their efficacy, an in vitro tissue-level model would be valuable. Nucleus pulposus (NP) explant culture may constitute such a model, as the earliest signs of degeneration are in the NP. However, in NP explant cultures, balancing tissue osmolarity is crucial to preventing swelling, proteoglycan (PG) loss and, therefore, maintaining a native cell environment. In this study, we investigated the effect of medium osmolarity on NP explants. We hypothesized that balancing the inherent tissue osmolarity would prevent swelling and thus maintain NP tissue in a native state. Bovine NP explants were cultured for 21 days in hypo-, iso-, and hyper-tonic conditions using either sucrose or polyethylene glycol (PEG) to raise medium osmolarity. Explants were analyzed for water and biochemical content, cell viability, gene expression, and tissue histology, and compared to day 0 samples. In hypo-tonic and both sucrose cultures, swelling was not prevented, resulting in PG loss and changes in cell behavior. Only PEG cultures maintained water and biochemical content and a histological aspect similar to those of native tissue, with better results for hyper- than for iso-tonic conditions. Using PEG to raise culture medium osmolarity, we were able to maintain the NP tissue specific matrix composition, important for disc cell behavior. This approach, thus, constitutes a promising model to test regenerative therapies for early intervertebral disc degeneration.