Balloon angioplasty for infant coarctation is associated with a high recurrence rate, making operative repair the gold standard for low-risk infants. Debate exists as to whether high-risk infants might be better served with primary angioplasty. We compared the outcome in high-risk versus low-risk infants over 20 years, in a center that always used surgical repair as the primary intervention. Of 192 infants from 1986 to 2005, 56 were considered "high-risk," defined as requiring prostaglandin infusion together with either epinephrine infusion for 24 hours preoperatively, or ventilation and milrinone infusion for 24 hours preoperatively. All high-risk patients had a period of ventricular dysfunction prior to surgery, ranging from mild to severe. Outcomes were compared using Bonferroni comparison of means or the Fischer exact test as appropriate. Although the high-risk patients were smaller (3.3 ± 0.1 vs 4.2 ± 0.2 kg, p < 0.01), younger (18 ± 4 vs 57 ± 7 days, p < 0.01), and more often required a concomitant pulmonary artery band (25% vs 15%, p = 0.05), their cross-clamp times were the same as the low-risk patients (18.9 ± 0.9 vs 18.0 ± 0.4 minutes, p = 0.27) and there was no difference in postoperative morbidity (7% vs 3%, p = 0.11). However, there was a trend toward higher perioperative mortality (7% vs 2%, p = 0.07). When compared with the published studies of primary angioplasty in comparable high-risk infants, the mortality rate in our surgically treated high-risk group is much lower. Additionally, only 11% of our high-risk group required reintervention, with two-thirds treated successfully with a single angioplasty at 3.8 ± 2.2 years later, far lower than recurrence rates with primary angioplasty. We propose that primary surgical repair of coarctation in infants who are high risk should be the primary treatment, with angioplasty reserved for recurrent coarctation.