Eighty-two consecutive patients with interrupted aortic arch were referred to our institution between 1985 and 1995. Three died before any attempt at operation and 79 underwent surgical repair. Median age at operation was 9 days (range 1 day to 6 years) and median weight was 3.0 kg (range 1.8 to 20 kg). All but one were in severe congestive heart failure and 31.5% had oliguria or anuria. Preoperative pH varied between 6.8 and 7.4 (median 7.3). Sixty-nine received prostaglandin E 1 infusion and 54 received mechanical ventilation. Aggressive preoperative ressucitation was necessary in 43 cases. Preoperative transfontanellar echography (performed routinely) since 1987 revealed intracerebral bleeding in six patients. Type A interrupted aortic arch was present in 37 cases, 41 patients had type B, and one had type C. Interrupted aortic arch was associated with single ventricular septal defect in 35 cases, 24 patients had associated complex heart defects, and 30 had significant subaortic stenosis (six had both subaortic stenosis and complex association). Aortopulmonary window was found in four patients, truncus arteriosus was found in eight, and transposition of the great arteries was found in five, double-outlet right ventricle was found in one, single ventricle was found in three, multiple ventricular septal defects were found in two and superior-inferior ventricles were found in one. Sixty-four patients underwent single-stage repair and 15 underwent multistage repair. Aortic arch repair consisted of direct anastomosis in 59 cases, patch augmentation in eight, and conduit interposition in 12. Ten patients underwent associated pulmonary artery banding and 19 underwent concomitant repair of complex associated lesions. The subaortic stenosis was addressed by four surgical techniques: myotomy or myectomy in five patients; creation of a double-outlet left ventricle, aortopulmonary anastomosis, and conduit insertion between the right ventricle and pulmonary artery bifurcation in four; no direct attempt to relieve the subaortic stenosis in six; and left-sided ventricular septal defect patch in 15. Mean duration of deep hypothermic circulatory arrest, crossclamp time, and cardiopulmonary bypass time were 38.8 ± 15.6 min, 60.5 ± 24.7 min, and 143 ± 40.1 min, respectively. Postoperative mortality rate was 18.9% (70% confidence limits 14% to 24.6%), and overal mortality rate was 31% (70% confidence limits 20.9% to 42.2%). The results have improved with time, with an overall operative mortality rate of 12% since 1990. Univariate statistical analysis revealed that early survival was influenced by preoperative renal function, detection of cerebral bleeding by transfontanellar echography, the number of cardioplegic injections, and the date of operation. Multivariate analysis revealed that preoperative renal function and the number of cardioplegic injections were independent risk factors for early mortality. Echocardiographic measurements of the left heart–aorta complex with preoperative Z values as low as -4 demonstrated rapid growth after repair. In the presence of subaortic stenosis, better survival was obtained with a left-sided patch for ventricular septal defect closure ( p < 0.05). Twenty-three patients underwent 26 reoperations for recoarctations (seven), left bronchial compression (two), second-stage repair (eight), right ventricle–pulmonary artery conduit replacement (three), and miscellaneous (four). One of the survivors was reoperated on for subaortic membrane. Survival at 5 years for the entire series was 70%. For isolated forms, it was 73.5% (90% for 1990 to 1995), for complex forms it was 70%, and in the presence of subaortic stenosis it was 60%. In conclusion, interrupted aortic arch remains a surgical challenge with continually improving results. Early diagnosis with preoperative resuscitation and adequate myocardial protection seem extremely important for further improvements. Associated subaortic stenosis or complex lesions can be treated in a single-stage repair with satisfactory medium-term results. (J T HORAC C ARDIOVASC S URG 1996;112:1150-60)