Purpose: To report the results of a two-center study of endovascular abdominal aortic aneurysm (AAA) exclusion using a polyester-covered nitinol stent-graft. Methods: Candidates were evaluated with arteriography and computed tomography. Criteria for endovascular therapy were a proximal aortic neck > 10 mm in length and < 25 mm in diameter, no bilateral internal iliac artery involvement in the aneurysm, no markedly tortuous common iliac arteries (CIAs) or CIAs < 7 mm in diameter, and no superior mesenteric artery occlusive disease. Patients were treated with the Mialhe Stentor and Vanguard stent-grafts in either tube or bifurcated versions. Results: Between August 1994 and November 1996, 149 patients (mean age 67 years, range 49 to 90) were admitted to the study. Overall primary technical success (aneurysm exclusion without endoleak) was 87% (130 patients): 78% (7 patients) for tube grafts and 88% (123 patients) for bifurcated endografts. The rate of local, remote, or systemic complications was 10.8%, with a 30-day mortality rate of 0.7%. During an average 13.5-month follow-up, there were no late deaths. Four of 20 endoleaks sealed spontaneously, 14 were treated with endoluminal techniques, and 2 remain untreated by patient request. Three graft limb thromboses occurred; one was treated surgically, one with lytic therapy, and one was untreated. Secondary patency was 96%. Conclusions: Endoluminal repair of infrarenal AAAs using straight or bifurcated grafts is a feasible alternative to conventional surgical repair. Longer follow-up and more experience with refined endograft models will elucidate the durability of this endovascular approach to treating AAAs.