English child protection register statistics reveal wide and persisten
t variations between authorities, but over time their rank order remai
ns moderately stable. Sociodemographic characteristics of areas combin
ed with differing policies and practices in keeping the register have
been found to explain much of the variability. While statistics of chi
ldren on registers are not valid measures of the incidence of child ma
ltreatment, registers themselves appear to play a useful part in local
child protection systems.