Standardized approaches to the treatment of neonatal seizures remain undeveloped. We assessed the type and number of anticonvulsants selected, blood levels attained, and postdischarge anticonvulsant treatment of neonatal seizures among five neonatal intensive care units in the United States between 2000-2003. Almost all of the 480 neonates (94%) with seizures were treated, initially with phenobarbital (82%), lorazepam (9%), phenytoin (2%), other anticonvulsants (1%), or a combination of the first two drugs (6%). While the majority of neonates were treated with one drug (59%), the number of anticonvulsants varied (P<0.0001), as did the peak serum phenobarbital levels (P<0.0001). The majority (75%) of survivors received anticonvulsant treatment after discharge. These neonates were more likely to have had abnormal electroencephalography or brain imaging, or to have needed a second anticonvulsant, compared with neonates whose drug therapy was discontinued. Anticonvulsant therapy is used in the majority of neonates with seizures, mostly with phenobarbital, and treatment is continued beyond discharge. The observed wide therapeutic variability may reflect a lack of standardized diagnostic and treatment approaches, particularly for seizures refractory to initial phenobarbital therapy. Trials of anticonvulsants with long-term neurodevelopmental follow-up are needed to develop evidence-based treatment guidelines.