An analysis has been made in 1080 aorto-iliac reconstructions performed from 1959 through 1974 at the Department of Surgery, University of Heidelberg. The ages of 1071 patients ranged from 20--97 with an average of 54.2 years. The ratio of men to women was 19 to 1. The predominant underlying lesion was atherosclerosis in 87.9 %. The choice of reconstructive procedure is dictated by the nature and extent of the occlusive process beside certain systemic factors. Morphologic characteristics allow the classification into 4 categories of aorto-iliac obliterations: Segmental occlusions (type I, 44 %), diffuse involvement of the iliac arteries and the terminal aorta (type II, 30 %); limitation of the lesion to the bifurcation (type III, 17%); propagation of the process towards the renal arteries (type IV, 9%). From 1959 through 1965 dacron bypass grafting was the most frequently employed procedure. Since 1961 a small number of endarterectomies was carried out. Since 1966 endarterectomy was used in a rising number of cases and has become the most common type of aorto-iliac recontruction in the last 4 years. The indication for endarterectomy is restricted to segmental occlusions (type I), to lesions of the bifurcation (type III) and in younger patients to diffuse obliterations of the iliac arteries and the terminal aorta (type II). The dacron bypass graft is the method of choice in the greater part of diffuse lesions (type II) and in aortic occlusions up to the renal arteries (type IV). Endarterectomy was performed in 478, bilateral bypass grafting in 530 and unilateral bypass grafting in 72 cases. Stage II, i.e. intermittent claudication, was the indication in 77%, stage III and IV, i.e. resting pain and gangrene, in 23% of all reconstructions. Associated occlusive disease of the femoropopliteal arteries was present in 62 %. The over-all operative mortality was 6.8%. For all the series endarterectomy showed a patency rate of 90.7%, bilateral bypass grafts of 92.2%, and unilateral bypass grafts of 77.2%. Accumulative patency rates by the life table method do not show statistically significant differences between endarterectomy and bilateral 42 U. Schulz et al. bypass grafts. Our preference for endarterectomy over bypass procedures in recent years seems also to be justified by the uncertain fate of the prostheses and their incorporation.