Relative motion between the brain and skull may explain many types of brain injury such as intracerebral hematomas due to bridging veins rupture and cerebral contusions. However, no experimental methods have been developed to measure the magnitude of this motion. Consequently, relative motion between the brain and skull predicted by analytical tools has never been validated. In this study, radio opaque markers were placed in the skull and neutral density markers were placed in the brain in two vertical columns in the occipitoparietal and temporoparietal regions. A bi-planar high-speed x-ray system was used to track the motion of these markers. Due to limitations in current technology to record the x-ray image on high-speed video cameras, only low speed (<4m/s) impact data were available. A previously developed finite element model of the brain simulating blunt head impact was used to study the feasibility of using this model to obtain relative displacement of the same magnitude as that obtained experimentally. The model simulated the scalp, three-layered skull, dura, falx, tentorium, pia, cerebral spinal fluid (CSF), venous sinuses, ventricles, cerebrum (gray and white matter), cerebellum, brain stem, and parasagittal bridging veins. No sliding was allowed between any component structures of the model, and a layer of solid elements with low shear modulus was used to model the CSF. However, this approach was not able to predict relative motions over 1 mm between the brain and skull. In this study, the model was modified. Although the CSF remained as a layer of material with a low shear modulus, a sliding interface was introduced to simulate the interaction between the CSF and pia matter. With this change in place, the model predictions corresponded with brain displacement data obtained from cadaveric experiments. The relative skull/brain displacement-time histories predicted by the new model agreed well with those obtained experimentally. The model was also run at higher impact speeds to obtain intracranial pressure as well as displacement histories. The computed coup/contrecoup pressures and contact forces predicted by the model compared favorably with the experimental data published by Nahum et al. Simulation results reproduced the translational acceleration injury mechanism (coup/contrecoup) proposed by Gurdjian and Lissner.