In order to examine neurological outcome in twins, a multicentre prosp
ective cohort study of infants was used. Women entered at the first pr
enatal visit, with follow-up of offspring to the age of 7 years. Outco
me measures were death, cerebral palsy, seizure disorders including ne
onatal, febrile or nonfebrile, and intelligence quotient. Among a tota
l of 52 364 livebirths, there were 1079 twins. Overall, rates of fetal
or neonatal death and cerebral palsy were higher in twins. Neonatal a
nd febrile seizures occurred with similar frequency in twins as in sin
gletons. Although twins were much more likely than singletons to be lo
w in birthweight, twins < 2500 g were not at higher risk for cerebral
palsy than low birthweight singletons. The risk of cerebral palsy and
of nonfebrile seizure disorders was similar in monozygous and dizygous
pairs, and in like-sex and unlike-sex pairs. Death of one twin was as
sociated with higher rates of cerebral palsy and of nonfebrile seizure
disorders in survivors, but not with lowered intelligence. Data from
this cohort suggest that low birthweight and survival status of the co
-twin are dominant predictors of childhood neurological morbidity in t
wins.