In a series of 18 polytetrafluoroethylene femoropopliteal bypasses performed for limb salvage, 33 mo. life-table cumulative patency rates for graft inserted into an isolated popliteal artery segment (91% .+-. 5%) were not significantly different from those inserted into a popliteal segment with angiographically better runoff (78% .+-. 5%). Below-knee reconstructions had 3 yr patency rates (86% .+-. 6%) that were not significantly different from those for bypasses inserted into the popliteal artery above the knee (79% .+-. 6%). One clear disadvantage of the isolated popliteal artery segment as a site for distal insertion of the bypass was the increased incidence of limb loss despite a patent reconstruction. This was particularly frequent in [human] diabetics with extensive foot gangrene or infection and could be avoided by a secondary extension of the bypass to a patent distal leg or foot artery.