The purpose of this study was to evaluate the clinical outcome of composite valve graft replacement in 193 patients with aortic valve disease and aneurysm of the ascending aorta from January 1980 to June 1999. The clinical outcome was compared between the patients diagnosed with Marfan syndrome (M group) and those without Marfan syndrome (non-M group), between those with aortic dissection (AD group) and without dissection (non-AD group), between 2 different techniques for coronary artery reattachment (modified Bentall [mB] and modified Piehler [mP]), and between the time of operation (1980-1989 and 1990-1999). Long-term outcome of this procedure was almost satisfactory with actuarial survival of 71.5 +/- 4.4% at 10 years and freedom from reoperation of 76.5 +/- 4.4% at 10 years. Freedom from cardiovascular events and freedom from reoperation were significantly lower in the M group and AD group than in the non-M and non-AD groups. Also, actuarial survival was significantly higher in the latter 10 years compared with the former 10 years. It was concluded that the improvement of perioperative management and proper selection of the technique for coronary artery reattachment could have improved the clinical outcome. In patients with Marfan syndrome or aortic dissection, there still remains a higher risk of cardiovascular event and future reoperation. Extensive aortic reconstruction or staged operation should be performed in such patients.