OBJECTIVE: Patients with type B aortic dissection differ from patients with type A dissection in age, hypertension prevalence, indications andtiming of surgical treatment, yet reported long-term results have been rather similar (see Doroghazi et al. J Am Coll Cardiol 1984;3:1026-1034). METHODS: With the aim of comparing the post-surgical history, we have reviewed our results in 288 dissections, 213 type A and 75 type B, operated consecutively between 1 January 1970 and 31 November 1994. Follow-up was 100% complete. Empirical survival of both groups was interpolated with a fully parametric method and the shape and scale of the hazard function was investigated. RESULTS: Survival was not significantly different betweentype A and type B. Parametric survival was, respectively, 0.52% (70% C.L.: 0.48-0.55) vs. 0.56% (0.51-0.62) at 5 years, 0.44% (0.40-0.47) vs. 0.28%(0.23-0.25) at 10 years, 0.37% (0.33-0.41) vs. 0.25% (0.19-0.32) at 15 years, and 0.31% (0.26-0.35) vs. 0.24% (0.18-0.31) at 20 years. Following the high perioperative risk phase in type A dissection, the intermediate and late risk remains constant at a rate of 0.0033 events/month (3.9% patient-years (pt.-years)). By contrast, the postoperative course of type B dissection shows an intermediate risk phase between 4 and 10 years withan average linearized risk of 9.3% pt.-years and a peak of 20%. This determined lower survival rates (24 vs. 31% at 20 years, P = NS). CONCLUSIONS: We conclude that patients with type B dissection have a steeper postoperative death hazard as compared to type A dissection patients. Age confounding or late entry do not explain the difference. This could be possibly related to a greater propensity for expansion, higherrisk of malperfusion complications or to limitations of our current surgical treatment.