Determination of human tolerance to injury is difficult because of the physical differences between humans and animals, dummies and cadaver tissue. Certain human volunteer testing has been done but at subinjurious levels [STAP 86] [EWIN 72] Considerable biomechanical engineering injury studies exist for the adult human cadaver however little is available for the pediatric population [SANC 99] [KLEI 98b]. Studies have been made of pediatric skull bone modulus, fetal tendon and early pediatric studies of the newborn during delivery, however, a paucity of information still exists in these areas. A number of dummies have recently been made available principally for airbag testing to bridge the gap between the 50 percentile Hybrid III male dummy and the 95 percentile male dummy. A review of the air bag dummies including the 12-month old CRABI dummy, 3-year old Hybrid III dummy, 6-year old Hybrid III dummy, the Hybrid III small female, Hybrid II mid-size male and 95 percentile large male is given. The major source of the biomechanical data for the development of the child dummies is direct biomechanical scaling. It can be shown that direct biomechanical scaling cannot represent the forces required for injury, because material properties are not included.Scaling for the neck using Nij for the four possible modes of injury have been developed. New criteria have been developed by Hertz for the probability of skull fracture based upon a log normal curve to replace the Prasad-Mertz Risk of Life Threatening Brain Injury curve. Rollover and air bag injuries are discussed.