During a 9 year period between January 1977 and December 1985, 98 consecutive infants under 3 months of age underwent surgical repair of symptomatic aortic coarctation. Resection and end-to-end anastomosis was performed in 73, subclavian flap angioplasty in 14, and other procedures in 11 patients. There were 20 (20.5%) early and 12 (12.5%) late deaths. No early deaths occurred in the isolated coarctation group. Associated complex cardiac malformations and age under 2 weeks at operation influenced significantly early and late outcome but not any particular surgical procedure. The survivors were followed from 6 months to 8 years and 8 months postoperatively. There were 16 (28%) re-coarctations among 56 survivors after end-to-end anastomosis requiring re-operation in 7 (12%) infants and 3 (30%) re-coarctations among 10 survivors after subclavian flap angioplasty requiring re-operation in 1 infant. After end-to-end anastomosis re-coarctation as well as re-operation rate was markedly lower when an interrupted suture line for the entire anastomosis was used as compared to the group with a continuous suture line of the posterior aortic wall (21% vs. 33% re-coarctation rate and 4% vs. 18% re-operation rate respectively). From our results it is concluded that subclavian flap angioplasty for relief of aortic coarctation in early infancy is not superior to resection and end-to-end anastomosis. In the end-to-end anastomosis group an interrupted suture line has a lower re-coarctation as well as re-operation rate as compared to a continuous suture line of the posterior aortic wall.