Traumatic brain and skull injuries are the most common serious injuries sustained by children in motor vehicle crashes. To address head protection for children, regulations typically use the Head Injury Criterion (HIC), scaled from adult and subhuman primate data and based on tolerance for skull fracture. It is unclear whether the spectrum of head injuries sustained by children in motor vehicle crashes is adequately addressed by a fracture-based criterion. To provide insight into this question, this paper evaluated patterns and predictors of pediatric head injuries using a large US population- based surveillance system and clinical data from a large pediatric trauma center. Predictors of elevated AIS2+ head injury include lap belt only (1.2%), right front seat position (1.4%), rollovers (4.6%) and struck-side crash (1.6%). Examination of non-concussive head injuries showed similar relationships. Increasing age was associated with increasing non-concussive head injury risk except for the children less than 1 year whose injury risk was slightly higher than their older counterparts. Review of the trauma registry data revealed that approximately 60% of the head injuries sustained by children age 1-12 years were intracranial injuries in absence of skull fractures. Consideration of the known mechanisms of the specific head injury diagnoses illustrated a potential gap in regulation by highlighting the substantial fraction of injuries that may be better correlated with a head injury metric derived from a measure of head rotation than the current HIC measurement.