Between January 1976 and March 1987, 78 patients underwent surgery for chronic aortic dissection at our institution. The ascending aorta was involved in 66 cases (Stanford type A) and was not involved in 12 cases (Stanford type B), wherever the initial dissection was suspected. Aortography remains the main preoperative investigation. The surgical technique varies according to the type of dissection. It seems essential to exclude the primary intimal tear and all dilated segments of the aorta must be replaced. The overall operative mortality was 11.5% (7.5% in type A, 33.3% in type B dissection). Sixty-three patients have been followed for a period varying between 6 months and 10 years (mean 5 years). The overall survival at 6 years is 60% +/- 5.6%. Because of the ultimate risk of aneurysmal dilatation of the false channel, these patients must be followed by CT scanning, colour flow Doppler echocardiography, magnetic resonance imaging, and in some cases, aortography.