Low back pain (LBP) was routinely subjectively assessed by Visual Analogue Scale (VAS) and Oswestry Disability Index (ODI) in clinical practice. However, the results of the assessment depend on the patients' subjective feeling. The aim of this study was to explore the quantitative parameters correlated to pain of the lumbar disc herniation (LDH) and nonspecific LBP (NLBP) subjects distinction with healthy control group (HG) during dynamic exercises using inertial sensor. Participants were asked to perform twelve maximum voluntary spine motions (forward bending (FB), backward bending (BB), left and right lateral bending (LLB and RLB), left- and right-axial rotation (LR and RR)) in standing position, and the corresponding motions (SFB/SBB/SLLB/SRLB/SLR/SRR) in sitting position without pain. Mean maximum angle, angular velocity and angle velocity of variation (ZL) were calculated. Results showed that there was high negative correlation between mean of maximum spine angle for subjects at each sublevel and VAS pain level/ODI score during eight movements (LR/RR/LLB/RLB/SLR/SRR/SLLB/SRLB). In addition, there was moderate correlation between the peak angular velocity/ ZL and VAS pain level for FB movement. There was a significant difference of maximum spine angle and angular velocity among three groups during LR/RR motions in sitting and standing position. These results indicate that the mean maximum angle and angular velocity of spine motions can provide a quantitative assessment during the rehabilitation progress of LBP patients and serve as supplementary method to distinct with LDH and healthy subjects in clinical practice.