A retrospective investigation was performed to study the benefit obtained from a vein cuff inserted at the distal anastomosis of a primary femoropopliteal bypass graft with polytetrafluoroethylene (PTFE). When PTFE was preferred to vein an attempt was made to assess whether there was subsequent use for the spared vein. A series of 559 primary PTFE bypasses was studied and found to have a cumulative patency rate of 62 per cent at 36 months. Subgroups were also studied for a similar period. In the above-knee position there was no difference in the primary patency rate of PTFE with or without a cuff at 36 months (69 versus 68 per cent, P = 0.89). Below the knee, grafts with a cuff had a better 36-month patency rate than those without (57 versus 29 per cent, P = 0.01). Cuffed PTFE for claudication had a better patency rate than that for threatened limbs at 36 months (68 versus 60 per cent, P = 0.03). Thrombectomy of occluded cuffed PTFE grafts improved the 36-month patency rate from 64 to 74 per cent. The limb salvage rate at 36 months was 97 per cent for claudication and 89 per cent for threatened limbs with cuffed PTFE. Occlusion of cuffed PTFE was associated with preservation of outflow in 91 (51 per cent) of 179 cases, permitting subsequent graft thrombectomy or insertion of a new bypass. Spared vein was rarely used to salvage an occluded PTFE graft, and was associated with poor patency and a high amputation rate. Although this study does not support the use of PTFE for the primary graft, PTFE with a vein cuff interposed at the distal anastomosis is a reasonable substitute when vein is not available.