When birth weight for gestation is used as a surrogate for intrauterine growth, the prevalence of cerebral palsy varies continuously in a reversed J shape, with steep increases in the risk for infants lighter and heavier than the optimum size. Patterns of size-at-birth specific risk for cerebral palsy differ between male and female infants, as do the patterns for more severe versus milder cases. Although these excess risks with abnormal size at birth imply antenatal precursors, it is not clear whether or how intrauterine growth is involved in any of the suspected causal pathways resulting in cerebral palsy. The implication for clinicians is that serial measures of in utero growth may provide an important indicator of fetal health.