Nucleus replacement has been proposed as a treatment to restore biomechanics and relieve pain in degenerate intervertebral discs (IVDs). Multiple nucleus replacement devices (NRDs) have been developed, however, none are currently used routinely in clinic. A better understanding of the interactions between NRDs and surrounding tissues may provide insight into the causes of implant failure and provide target properties for future NRD designs. The aim of this study was to non-invasively quantify 3D strains within the IVD through three stages of nucleus replacement surgery: intact, post-nuclectomy, and post-treatment. Digital volume correlation (DVC) combined with 9.4T MRI was used to measure strains in seven human cadaveric specimens (42 ± 18 years) when axially compressed to 1 kN. Nucleus material was removed from each specimen creating a cavity that was filled with a hydrogel-based NRD. Nucleus removal led to loss of disc height (12.6 ± 4.4%, = 0.004) which was restored post-treatment (within 5.3 ± 3.1% of the intact state, > 0.05). Nuclectomy led to increased circumferential strains in the lateral annulus region compared to the intact state (-4.0 ± 3.4% vs. 1.7 ± 6.0%, = 0.013), and increased maximum shear strains in the posterior annulus region (14.6 ± 1.7% vs. 19.4 ± 2.6%, = 0.021). In both cases, the NRD was able to restore these strain values to their intact levels ( ≥ 0.192). The ability of the NRD to restore IVD biomechanics and some strain types to intact state levels supports nucleus replacement surgery as a viable treatment option. The DVC-MRI method used in the present study could serve as a useful tool to assess future NRD designs to help improve performance in future clinical trials.