Improvements in obstetric care have certainly led to a decrease in the incidence of birth related spinal cord trauma but unfortunately it is still encountered from time to time. Its exact incidence is difficult to determine because the spinal cord is not routinely examined in all necropsies. Yates (1959) found evidence of distortional trauma to the cervical spine in 45% of perinatal necropsies.' Towbin (1969) reported significant spinal or brain stem injury in over 10% of neonatal necropsies.2 More recently Reid in 1983 reported 48 perinatal necropsies and concluded that although there was no decrease in the incidence of cervical spine trauma in comparison to Yates's series, there was a decrease in the severity of damage in individual cases.3 The incidence of serious spinal cord damage in our region over the last five years is estimated to be approximately one in 29 000 live births. Presently, the most common scenario for birth related spinal cord injury is after Kielland's forceps rotation for persistent occipitoposterior position of the presenting vertex4 (personal experience) in contrast to previous reports where 75% of the cases occurred after breech presentation.5 Other predisposing factors include hyperextension of the fetal head,6 external cephalic version,7 inherent congenital vertebral anomalies,8 prematurity,9 precipitate delivery,9 intrauterine asphyxia,9 10 benign congenital hypotonia,'0 and forceps application in presence of oligohydramnios, uterine contraction, or a constriction ring.'1 Occasionally spinal cord injury has occurred after an easy, uneventful normal cephalic delivery. 12 13 Clinical suspicion of birth related spinal cord injury is often delayed. Diagnosis and management are difficult and there are long term serious medical, legal, and ethical implications. Accurate localisation and assessment of the severity of the lesion is of vital importance in planning the management strategies. There is no single test which determines the site and extent of the lesion in every case and there is considerable discordance in the results of various neurodiagnostic imaging techniques and also among the different electrophysiological tests.