A subset of children with minor head injury is known to develop serious neurological consequences, but identifying this subset has been difficult. In adults, multiple variables such as skull fracture, Glasgow Coma Scale score, unconsciousness, and amnesia are significant factors that determine whether to admit the patient to the hospital and the patient's outcome. As an objective assessment of neurological function, however, the Glasgow Coma Scale has limited usefulness in children, particularly those under 36 months of age. We report our experience with 937 children having head injuries using a Glasgow Coma Scale modified for children (Children's Coma Scale). During the 6-year study period, 791 of the 937 children (84%) sustained minor head injuries (Glasgow Coma Scale or Children's Coma Scale score of 13, 14, or 15). The mean age of patients was 5.5 years. Males predominated over females by a 2:1 ratio. The most common cause of injury was a fall, followed by a pedestrian/motor vehicle accident. Seven hundred and thirty-nine of the 791 children were alert at the time of admission. Of these, 99 (13.4%) had lesions requiring surgery: 9 had subdural hematomas, 35 had epidural hematomas, 44 had depressed skull fractures, and 11 had other types of lesions. Two children (0.3%) with Glasgow Coma scores of 13 died after subsequent deterioration, 1 of a delayed epidural hematoma, the other of diffuse cerebral edema. Risk factors such as length of unconsciousness, presence of skull fractures, computed tomography findings, posttraumatic seizure, and Glasgow or Children's Coma scores were evaluated for their impact on the patient's outcome.(ABSTRACT TRUNCATED AT 250 WORDS)