Several important points for obtaining high-quality reliable visual evoked potentials (VEPs) in newborns while minimizing variability and the risk of misleading prognoses are presented.