The study aim was to compare the results of aortic valve repair and replacement with biological valves in adult patients with aortic insufficiency (AI) caused by congenital bicuspid aortic valve (BAV) METHODS: Forty-four patients who had aortic valve repair were matched for age and left ventricular function to 44 patients who had aortic valve replacement (AVR) with biological valves. Patients were followed annually using echocardiography. The mean follow up was 2.6 +/- 2.1 years for the repair group, and 3.5 +/- 2.1 years for the replacement group. Follow up was complete.There was no operative or late death in either group. Early postoperative echocardiography showed trace or no AI in 35 patients and mild AI in nine who had repair, and trace or no AI in 38 patients and mild AI in five who had AVR. The mean peak systolic gradient was 16.2 +/- 7.6 mmHg for repair and 13.2 +/- 7.2 mmHg for AVR. Four patients who had valve repair and two who had AVR, needed repeat aortic valve surgery because of progressive AI or endocarditis. Freedom from reoperation at five years was 91 +/- 5% for repair and 94 +/- 6% for replacement (p = 0.2), while freedom from moderate or severe AI at five years was 79 +/- 8% for repair and 94 +/- 6% for replacement (p = 0.024). The peak systolic gradient at follow up was 11.7 +/- 6.8 mmHg after repair and 13.3 +/- 9.6 mmHg after AVR (p = 0.4). There were no thromboembolic complications in either group.Repair of BAV is feasible in certain patients with AI, but the hemodynamics and clinical outcomes do not appear to be superior to AVR with biological valves during the first five years of follow up.