Differences in matrix compositions in human nucleus pulposus (NP) clinical samples demand different cell isolation protocols for optimal results but there is no clear guide about this to date. Sub-optimal protocols may result in low cell yield, limited reliability of results or even failure of experiments. Cell yield, viability and attachment of cells isolated from bovine NP tissue with different protocols were estimated by cell counting, Trypan blue staining and cell culturing respectively. RNA was extracted from isolated cells and quantified by Nanodrop spectrometry and RT-qPCR. Higher collagenase concentration, longer digestion duration and pronase pre-treatment increased the cell yield. Cell viability remained high (<5% dead cells) even after 0.2% collagenase treatment for overnight. NP cells remained to have high ACAN, COL2A1, CDH2, KRT18, and KRT19 expression compared to muscle cells for different cell isolation conditions tested. Digestion by collagenase alone without the use of pronase could isolate cells from human degenerated NP tissue but clusters of cells were observed. We suggest the use of the disappearance of tissue as an indirect measure of cells released. This study provides a guide for researchers to decide the parameters involved in NP cell isolation for optimal outcome. © 2015 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 33:1743–1755, 2015.