Despite the immaturity of the newborn brain, the neonatal period is reported to have a very high frequency of seizures. This review concludes that many of the neonatal events that are called seizures probably originate from subcortical structures, have little in common with cortical seizures seen in older individuals, and may not benefit from conventional anticonvulsant treatment. Many studies of anticonvulsants in the newborn have important methodologic problems, compounded by the fact that the seizures tend to spontaneously remit with the resolution of the acute hypoxic-ischemic encephalopathy that is most often the cause. Randomized trials of anticonvulsants in this setting have not been carried out. Even in many of these seizures do not originate in the cortex, they still imply profound cortical disturbance and are associated with high mortality and morbidity. It is unknown if the type and duration of treatment influence the long-term, overall outcome. The seizures usually stop in the newborn period, and anticonvulsants beyond hospital discharge seem unwarranted because they are unlikely to prevent subsequent epilepsy. Newer investigations, including video-EEG and nuclear magnetic resonance studies, may clarify the real significance of neonatal seizures.