The purpose of the study was to look at how nurses and physicians of a university hospital rated their ability to identify child maltreatment while caring for those children. In this study, child maltreatment was defined as physical, psychological or sexual abuse or neglect of a child under the age of 18 by parents or caregivers. The total population of staff caring for children (N=513) in a university hospital were surveyed. Data were collected with a questionnaire developed for this study with reference to the literature. Altogether 317 questionnaires were returned, which yielded a response rate of 62%. The data were analysed using statistical methods and quantitative content analysis. Forty per cent of the respondents estimated that they had never cared for a maltreated child. Two-thirds of the respondents believed that they would be able to identify a child maltreatment case. The most distinct signs by which maltreatment could be identified were fractures, multiple bruises and the fact that the child had frequent injuries. The child's or parent's behaviour often aroused suspicion of maltreatment. Seventy-one per cent of the respondents rated the identification of maltreatment as rather difficult or difficult. Awkwardness of the phenomenon, the staff's pressure of work and relative unfamiliarity with the phenomenon were assessed as things that make the identification difficult. The fact that no jointly agreed guidelines were available for handling the matter was seen as a particular weakness. The respondents had fairly much theoretical knowledge about child maltreatment. However, child maltreatment is a multi-dimensional phenomenon that calls forth emotions. The development and improvement of practical nursing and medical care and of staff collaboration require that education be provided to different occupational groups and parties caring for children and that jointly agreed hospital-specific and regional models for operation be developed.