During the 2-year study (1993 and 1994), 28 infants presented to the accident and emergency department with fractures; there were 17 boys and 11 girls, and their ages ranged from 1 month to 11 months, with the majority being older then 6 months. A clear history of trauma was evident in 24 cases, but in 4 cases, the explanation of the injuries was not consistent with the findings of the clinical exam. These four and one other case without a history of trauma were suspicious. Three of the cases had other clinical evidence of abuse. All five cases were referred to the pediatric service, where they were given skeletal surveys; three were found to have other fractures. All five cases were eventually diagnosed as nonaccidental fractures and were referred to the Social Services Department. In follow-up, home visitors involved with the families of the 28 children were surveyed to determine whether there had been any child protection services offered to the children and their families. A total of 25 questionnaires were completed. The responses indicated that the home visitors had more concerns about behavior and developmental issues in the children who had diagnosed nonaccidental fractures in infancy than in those who had nonaccidental fractures. This study confirms that abuse is a relatively common cause of fractures in infants. This requires follow-up for all infants presenting with fractures, given the difficulty of detecting which fractures are and are not caused by abuse. 8 references