Tolerance data for the spine are based on failure loads in mechanical tests of cadaveric spine specimens. These failure loads are often determined using changes in slope of the force or moment response curves, a method prone to subjectivity. Acoustic emission (AE) sensors may provide more objective data on the timing of injury; these have been used to detect the time of injury of facial and ankle bones but AE signals have not been reported for spinal ligaments tested at any loading rate or for any component of the spine in dynamic loading. The objectives of this study were to 1) compare the time of injury, as determined using AE signals, to those using traditional methods (time of injury observed in high speed video for ligament specimens and time of peak load observed on a force vs. time plot for bone specimens) and 2) compare the AE signals from vertebral body and ligament specimens in terms of amplitude and frequency. Isolated cadaveric vertebral bodies (VB, n=3) and ligamentum flavum (LF, n=3) specimens from the thoracic spine were tested in compression and tension, respectively, using a servohydraulic material testing machine while collecting AE signals and high speed video at 33,057 frames per second. Time of injury was determined using the peak amplitude of the AE signals and this was compared to that determined using traditional methods (time of peak force and time of first visual evidence of injury). AE signals from VB and LF specimens were also compared in terms of peak amplitudes and frequency contents. Two VB specimens failed by fracture (the mechanism of failure could not be visualized for one specimen) and the mechanisms of failure for the LF specimens were stripping of the ligament and periosteum from the bone (laminae). Time of injury determined using AE signals produced median differences (from those defined using traditional methods) of 0.5 and 0.2 ms for VB and LF specimens, respectively. VB fractures were associated with higher amplitude and higher frequency AE signals than those associated with LF failures (median amplitude 88.3 vs. 76.7 dB and median characteristic frequency of 55 vs. 27 kHz); although in this study this difference could not be evaluated statistically. Using AE signals, identification of the time of injury and differentiation between failures of different spine components was possible. Using these sensors it may become possible to decode complex failures of the spine that involve combinations of osseous and ligamentous failure that occur at different times.