Intervertebral range of motion (ROM) is commonly calculated using ordered rotations or projection angles. Ordered rotations are sequence-dependent, and projection angles are dependent upon on which orientation vectors are projected. This study assessed the effect of calculation method on intervertebral ROM in the subaxial cervical spine (C3–C7) during in vivo dynamic, three-dimensional, functional movement. Biplane radiographs were collected at 30 images per second while 29 participants performed full ROM flexion/extension, axial rotation and lateral bending movements of their cervical spine. In vivo bone motion was tracked with sub-millimeter accuracy using a validated volumetric model-based tracking technique. Intervertebral rotations were calculated using six Cardan angle sequences and two projection angle combinations. Within-subject comparisons revealed significant differences in intervertebral ROM among calculation methods (all p<0.002). Group mean ROM differences were small, but significantly different among calculation methods (p<0.001). A resampling technique demonstrated that as group size increases, the differences between calculation methods decreases substantially. It is concluded that the method used to calculate intervertebral rotations of the sub-axial cervical spine can significantly affect within-subject and between group comparisons of intervertebral ROM.