Head circumference (HC) is routinely measured in infancy and is widely assumed to be valuable in helping identify children with intracranial pathology or neurodevelopmental problems, and for monitoring such children over time. However the evidence base for screening HC is very limited. Very little is known about the extent to which deviation in head size or growth predicts underlying pathology. A common clinical assessment made where a child has a very large or small head is to compare the child's head circumference centile with that of their parents. However we do not at present know the extent to which children may vary from their parents in head size and we do know how well the existing growth reference (UK1990) age 20 fits to the normal distribution of adult heads. The ALSPAC data set provides a large population data set which will both allow us to explore the range of normality and supply enough numbers at the extremes to examine the extent to which extreme values do predict pathology. We have previously used the ALSPAC dataset to investigate the fit of UK children to the new WHO standard and found that ALSPAC children have larger heads than the WHO standard, and commonly cross centiles upwards. We will thus seek to answer two related research questions: 1/ How does parental head circumference correlate with head circumference centile in infancy and childhood, and what is the normal range of variance from this? (Concept: Parent-child HC) 2/ To what extent do variations in head circumference in the preschool years predict neurodevelopmental problems in childhood? (Concept: predictive value of HC). Head circumference (HC) is routinely measured in infancy and is widely assumed to be valuable in helping identify children with intracranial pathology or neurodevelopmental problems, and for monitoring such children over time. However the evidence base for screening HC is very limited. Very little is known about the extent to which deviation in head size or growth predicts underlying pathology. A common clinical assessment made where a child has a very large or small head is to compare the child's head circumference centile with that of their parents. However we do not at present know the extent to which children may vary from their parents in head size and we do know how well the existing growth reference (UK1990) age 20 fits to the normal distribution of adult heads. The ALSPAC data set provides a large population data set which will both allow us to explore the range of normality and supply enough numbers at the extremes to examine the extent to which extreme values do predict pathology. We have previously used the ALSPAC dataset to investigate the fit of UK children to the new WHO standard and found that ALSPAC children have larger heads than the WHO standard, and commonly cross centiles upwards. We will thus seek to answer two related research questions: 1/ How does parental head circumference correlate with head circumference centile in infancy and childhood, and what is the normal range of variance from this? (Concept: Parent-child HC) 2/ To what extent do variations in head circumference in the preschool years predict neurodevelopmental problems in childhood? (Concept: predictive value of HC).