Regenerative medicine approaches aiming at treating degenerating intervertebral discs, a major cause of back pain, are increasingly tested in ex-vivo disc explant models mimicking in-vivo conditions. For assessing the efficacy of regenerative therapies, cell viability is commonly measured requiring specific labels to stain cells. Here, we demonstrate and evaluate how cellular auto-fluorescence can be utilized to non-invasively assess viability in disc tissue in-situ using label-free two-photon microscopy. Live and dead bovine disc cells (0% and 100% cell viability) from the nucleus pulposus were seeded into collagen gels and auto-fluorescence was characterized. Subsequently, nucleus pulposus explants were cultured for 6 days in media with different glucose supplementation (0, 0.25, 0.5, and 1 g/L) to induce different degrees of cell death. Then, samples were split and viability was assessed using label-free two-photon microscopy and conventional staining. Results show that live and dead nucleus pulposus cells systematically emit auto-fluorescent light with distinct characteristics. Cell viability values obtained with label-free microscopy did not significantly differ from those acquired with staining. In summary, monitoring auto-fluorescence facilitates accurate cell viability assessment in nucleus tissue requiring no additional dyes. Thus, this technique may be suitable for pre-clinical testing of regenerative therapies in nucleus pulposus cultures. © 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 32:545-550, 2014. © 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.