Purpose In 1977, the use of gelatine-resorcine-formaline (GRF) biological glue during surgery of acute type A aortic dissection was proposed. The present study retrospectively analyses the late results obtained with this adjunct in an experience extending over a 20-year period. Patients and methods From January 1977 to July 1997, 193 patients (139 males and 54 females) aged from 15 to 79 years (mean age: 53 ± 14 years) underwent an emergency operation for type A aortic dissection in our institution. All patients suffering from acute type A dissection and 162 (84%) were operated on within 48 hours after the onset of symptoms. Twenty-eight patients (15.2%) had Marfan's syndrome. In all patients the ascending aorta was replaced and the aortic stumps were reinforced with the GRF glue. In 43 patients (22.2%), the aortic valve was replaced either independently (5 cases — 2.5%) or by means of a composite graft (35 cases −19.5%). Recently three patients underwent a complete replacement of the ascending aorta and coronary reimplantation with preservation of the native aortic valve. Because of the location of the intimal tear, the aortic replacement was extended to the transverse arch in 58 patients (30%). Results Hospital mortality amounted to 21% (40 patients) (22.8% in patients with arch replacement and 20.3% in patients without arch replacement) (ns). The survivors were surveyed from 2 months to 20 years post-operatively (cumulative follow-up: 856 pt/years, mean follow-up: 85 ± 66 months). During this period of time, 23 patients (15%) had to be reoperated on for a total of 29 procedures. Six of those patients (26%) died at reoperation. At univariate analysis, presence of Marfan's syndrome ( P < 0.05) and absence of arch replacement ( P < 0.02) were determinant risk factors for reoperation. Emergency ( P < 0.01) and thoraco-abdominal replacement ( P< 0.04) were determinant risk-factors of death at reoperation. The actuarial freedom from reoperation (Kaplan-Meier, Cl: 95%) was: 96.5% (90.9–98.2), 87.6% (79.8–92.7), 80.9% (70.8–88.1), 66.4% (51.1–78.9) at one, 5, 10 and 15 years, respectively. A total of 36 patients (27.7%) died during follow-up. Presence of Marfan's syndrome ( P < 0.01), reoperation ( P < 0.02), stroke ( P < 0.05), cardiac failure ( P < 0.05) were determinant risk factors of late mortality. The actuarial late survival rate (Kaplan-Meier. Cl: 95%), including hospital mortality, was: 71.5% (64.3–77.8), 66% (58.3–73), 56.4% (47.7–64.7), 46.3% (36.4–56.5) at one, 5,10 and 15 years. Conclusion The GRF glue has proved to be extremely useful during initial emergency surgery for acute type A dissection, making the procedure much easier and safer. Through this operative improvement, the use of the GRF glue seems to have a beneficial influence on the late results which, however, depend mainly on the patient's basic condition.