Purpose: The outcome of endoluminal repair of abdominal aortic aneurysms with two generations of prostheses was analyzed and compared. Methods: Between May 1992 and December 1998, 266 patients underwent elective endoluminal repair of an AAA. First-generation prostheses were used in 118 patients (group I), and second-generation prostheses were used in 148 patients (group II). The two groups were similar in age, sex, and size of AAA. The proportion of patients with comorbidities was higher in group I than in group II, but not significantly. First-generation devices were characterized by large (24F internal diameter) delivery systems, one-piece construction, and a lack of metallic support throughout their length. Second- generation devices had smaller (21F or smaller internal diameter) delivery systems, modular construction, and a metallic frame throughout the length of the prostheses. The major endpoint criteria were survival and successful endoluminal repair. Success was defined as the exclusion of the aneurysm sac from the circulation, with stability or the reduction in AAA maximum transverse diameter. Persistent endoleaks were classified as failures, irrespective of whether they were subsequently corrected by means of secondary endovascular intervention. The minimum follow-up period was 5 months for each of the 266 patients. Analysis was performed by means of the life-table method. Results: Perioperative mortality was not significantly different between group I (4.2%) and group II (2.7%). There was a statistically significant difference between the survival curves of the two generations, which favored group II (P = .012). There was a significant (P < .001) difference between the two generations of patients in their conditional probability of graft failure when the competing risk of all-cause mortality was considered. Second-generation patients were at a lower risk of graft failure than first-generation patients. The probability of failure, expressed as a proportion of grafts failing at 2 years, was 0.15 for patients with second-generation prostheses and 0.33 for patients with first-generation prostheses. Conclusion: Endoluminal AAA repair is a safe procedure, whether first- or second-generation prostheses are used. Survival and probability of graft success were significantly higher with second-generation prostheses than with first-generation prostheses. This improvement in outcome resulted from a combination of increasing clinical experience and advances in technology. A more accurate evaluation of the role of the endoluminal method in AAA repair would be achieved by studying patients in whom second- generation devices were used, rather than longer term studies in which first- and second-generation devices were used.