Rollover crashes are a major public health concern, associated with over one-third of all passenger vehicle fatalities in the US. Of the injuries sustained by occupants in rollover crashes, cervical spine trauma is among the most frequent and life threatening. Occupant initial orientation and loading distribution has been shown to influence injury outcome in biomechanical testing with cadaveric subjects; however, these loading conditions have not been well characterized for rollover-involved occupants. Slight changes in boundary condition or eccentricity in axial compression (the loading mechanism most associated with cervical spine injury in this crash mode) have been shown to have enormous bearing on the injury tolerance and severity of injury. A thorough review of cadaveric injury mechanism literature and of clinically representative rollover cases (queried from a national trauma database) was conducted for the purpose of retrospectively linking injury with failure mechanism and reverse-engineering the loading environments experienced by rollover-involved occupants with cervical spine trauma. Data gathered from these analytical observations, such as the influence of laterally eccentric loading, passive musculature, and torso augmentation, help establish an in vitro test approach relevant to injurious rollover-type loading. Cervical spine compression tests with full post-mortem human surrogates were performed at the Center for Applied Biomechanics with the goals of assessing 1) boundary condition influence on bony fracture in a full cadaver, 2) the effective mass of the human torso as it augments load to the base of the cervical column (the process believed by experts to be the greatest contributing factor to cervical spine injury in v rollovers), and 3) the subsequent buckling kinematics of the cervical spine during the initiation of an applied axial load. Dynamic high-speed x-ray was employed for this study, showing buckling of the spine within the first 12 milliseconds of axial loading, a phenomenon never before observed radiologically in situ. Bony fracture was produced in 3 of the 4 surrogates, all of which coincide with relevant rollover type fracture and compressive injury mechanism. Injury outcome differences, however, raise questions pertaining to the end conditions and effective torso mass employed by previous authors in simplified cervical spine studies, the same studies used for injury reference values in current neck injury criteria. Kinematics data was ascertained to determine the amount of cephalocaudal translation that is required to initiate 2nd-order buckling of the neck, a value found to be lower than previously believed. Findings of this study can be used in the design of component level (head-neck complex) biomechanical tests of the human cervical spine, a significant experimentation cost reduction from full-scale cadaveric testing. This study illustrates the disparity between previous test approaches and this author's methodology which lays the groundwork for future studies in determining ...