The purpose of the present study was to investigate whether an energy failure level applies to the skull fracture mechanics in unembalmed post-mortem human heads under dynamic frontal loading conditions. A double-pendulum model was used to conduct frontal impact tests on specimens from 18 unembalmed post-mortem human subjects. The specimens were isolated at the occipital condyle level, and pre-test computed tomography images were obtained. The specimens were rigidly attached to an aluminum pendulum in an upside down position and obtained a single degree of freedom, allowing motion in the plane of impact. A steel pendulum delivered the impact and was fitted with a flat-surfaced, cylindrical aluminum impactor, which distributed the load to a force sensor. The relative displacement between the two pendulums was used as a measure for the deformation of the specimen in the plane of impact. Three impact velocity conditions were created: low (3.60+/-0.23 m/sec), intermediate (5.21+/-0.04 m/sec), and high (6.95+/-0.04 m/sec) velocity. Computed tomography and dissection techniques were used to detect pathology. If no fracture was detected, repeated tests on the same specimen were performed with higher impact energy until fracture occurred. Peak force, displacement and energy variables were used to describe the biomechanics. Our data suggests the existence of an energy failure level in the range of 22-24 J for dynamic frontal loading of an intact unembalmed head, allowed to move with one degree of freedom. Further experiments, however, are necessary to confirm that this is a definitive energy criterion for skull fracture following impact.