A case-control study was carried out in Nottingham Health District, to establish whether children under five years of age admitted to hospital after an accidental injury were more likely to have previously attended the accident and emergency (A & E) department than community controls. The subjects were 342 case-control pairs matched on sex and date of birth, consisting of children under five years resident in the Health District, and the main exposure measures were attendance at the A & E department before the case's first admission, type of injury and number of earlier attendances. It was found that, after adjusting for social deprivation score and proximity to hospital, children who had been admitted after an accidental injury were twice as likely to have attended the A & E department than community controls, and were more likely to have had more than one earlier attendance. Odds ratios were significantly raised for softtissue injuries and lacerations. It is concluded that accidental injuries in pre-school children that require attendance at the A & E department predict accidental injuries requiring admission. Making attendances at A & E departments notifiable to health visitors would facilitate the undertaking of accident prevention work.