The death of a child through abuse or neglect is an event that touches all those concerned, and that may have effects which reverberate for a long time afterwards. For the parents the sense of loss is not necessarily less because they are responsible for the child’s death. For the professionals who were in direct contact with the child, their sense of loss and sadness may be mixed with extreme anxiety and stress in relation to their own careers, with the anticipation of censure both from the media and from their employing agencies. In one local authority, even before the preliminary facts were in, the Chair of the social services committee on learning through the media of the death of a child through abuse where the social services department had been involved with the family reacted angrily and perhaps injudiciously with the comment ‘Heads will roir. For staff of all agencies there may be concern over whether the agency will be considered to be at fault, and again the anticipation of strong adverse reaction from the media. Death is by no means the only risk in child protection; there are the risks of severe post abuse consequences for the child, the risks to children and families associated with child protection intervention, the risks to individuals other than the child. Nevertheless the particular risk of fatality has come to dominate the child protection system to the extent that it is often not balanced against other risks. Instead the system has been driven towards zero tolerance for the risk of fatality. Pritchard (1993) for example states, ‘the most crucial issue for child protection services is how successful or otherwise they are in their ability to prevent the deaths of children’. In that sense the system could be described as ‘death avoidant’. In the terminology of risk theory, child abuse fatality is a high consequence, low frequency risk which in the construction of child protection systems has not been balanced against high frequency, ‘low’ consequence risks. We have discussed elsewhere the impact of perception of risk on child protection systems (Sanders et al, 1996c; Jackson et al, 1995). The role of risk in the social construction of child abuse has been considered by Parton and colleagues (Parton, 1985; Parton et al, 1997). By arguing against a positivist or medico-social model of abuse they have emphasised the importance of definitions and thresholds, reflecting some of the work of the Dartington team in their overview of the Department of Health commissioned reports (Department of Health, 1995). Parton (1996) in a critique of that work notes that it fails to fully take on board the signficance of risk, which in the policy context of contemporary child protection has 112taken on a heightened signficance because of the tension between the need for greater targeting arising from diminishing resources.