In the surgery of aortic dissection, only a small section of aorta compared to the extent of the aortic damage, is usually replaced. The disease is not cured by surgery and needs continuous postoperative surveillance and medical therapy. We report the follow-up of 105 patients who were operated upon between January 1970 and April 1986 and discharged from hospital. Overall actuarial survival was 90% at 5 years, 52% at 10 years and 39% at 15 years. There were 20 deaths, mostly (85%) related to cardiovascular causes. Survival times were correlated, using multivariate methods, with several pre-, peri- and postoperative variables to identify significant risk factors and to calculate actuarial survivals. We found that postoperative low output syndrome (p = 0.007) and stroke (p = 0.04) adversely affected survival and that previous aortic disease or operation (p = 0.004) was associated with an increased rate of dissection-related complications. On the contrary, preoperative hypertension was related to a significantly better survival (p = 0.01) and survival free of dissection-related complications (p = 0.001). When dissection was related to hypertension, adequate postoperative medical treatment neutralized the progression of the aortic damage and its consequences. When dissection was not due to hypertension, the observed survival was unsatisfactory, probably because of a more fragile aorta and inadequate medical follow-up therapy.