Since there is no consensus as to the choice of an optimal prosthesis for aortobifemoral bypass, we have retrospectively evaluated the outcome with 162 Milliknit prostheses laid in cases of obliterating atherosclerosis. Such prostheses were preferentially used when the distal anastomosis were to be located on the deep femoral arteries. Intraoperatively, we noted the absence of fraying and the easy manipulation, owing to the flexibility and thinness of the wall, which are similar to those of the deep femoral artery: however, the blood loss before achieving complete tightness was greater than with most other knitted prostheses, and in 7.4% of the operated patients, defribination on clamp removal caused an additional bleeding averaging 1 litre. There was no graft sepsis during the postoperative period, may be because of the excellent healing of this prosthesis. One prosthesis only presented with excess dilatation accounting for thrombosis. The average length of follow-up was 3.6 years, the primary actuarial patency at 1, 5 and 10 years respectively was 98.7, 93.9 and 92.8%. Secondary patency at the same times was 99, 97.5 and 95.9%. These patency rates are among the best published in the literature. As a conclusion, this prosthesis is perfectly appropriate when the shunts must be located on thin femoral arteries, but a reduction of the intraoperative blood loss through an impregnation process would be desirable.