Ninety-nine patients underwent surgery for thoraco-abdominal (n = 73) or suprarenal (n = 26) aortic aneurysm, between January 1, 1981 and May 10, 1988. The in-graft technique was combined with re-implantation of renal, visceral, and segmental arteries, using the method proposed by Crawford. Ischaemic tolerance of kidneys was extended by means of flush cooling and was more recently manipulated by means of prostaglandin E1. Spinal cord function had been monitored since 1985 by means of the authors' method of spino-electrogram reading, with intercostal arteries being re-implanted in cases of need. The intervention was survived by 79 patients, with ruptures in 69 per cent of all cases and no ruptures in 82 per cent. The rate of paraplegia amounted to seven per cent (with 2.5 per cent of survivors. 25 per cent of deaths, 18 per cent with dissecting aneurysm, and five per cent without dissection), and it depended on the extent of aortic replacement (Types I, II, and III for 773 patients and Types IV and V for 0/26). While the incidence of postoperative disorders of renal function, including temporary dialysis, was also somewhat related to the extent of aortic replacement (Types I, II, and III 22 per cent and Types IV and V eight per cent), it was clearly affected to a greater extent by the general condition of patients (ruptures in 50 per cent, intact and without dissection in nine per cent). Spinal monitoring and protection against ischaemia have substantially contributed to much better surgical results, in recent years. Surgical treatment, therefore, is recommended for patients with aneurysmol symptoms as well as for advanced cases of aneurysm, the more as rupture-related mortality has proved to be extremely high in spontaneous courses without surgical action.