The outcome of 40 consecutive newborns and children with normal cardiac connections presenting with unoperated (i.e., no previous surgical or balloon dilation treatment) aortic coarctation since the introduction of balloon dilation as a treatment option (group A) was compared retrospectively with that of a similar cohort of 38 patients (group B) presenting in the period immediately before the introduction of balloon dilation. Group A consisted of 18 newborns (<30 days) and 22 children and Group B consisted of 20 newborns and 18 children. In group A, 30 patients had balloon dilation but 10 underwent surgery because of parental preference, unfavorable anatomy or for logistic reasons. Group B had conventional surgical repair by various methods. All patients were followed up clinically by Doppler echocardiography and, when possible, by magnetic resonance imaging. Repeat cardiac catheterization was only performed in those with suspected recoarctation. Survival and freedom from reintervention were investigated using Kaplan-Meier analysis. Patients of all ages in group A and those who had balloon dilation were at higher risk for reintervention (25 and 30%, respectively, at 1 year) than those in group B or those who had surgery (20 and 18%, respectively, at 1 year). Mortality was not significantly different between groups A and B, but there were 2 neonatal deaths directly attributable to the dilation procedure. Two patients developed aneurysms after dilation. Balloon dilation is an effective treatment for unoperated coarctation, but the recoarctation rate appears to be higher than for surgery.