Between January, 1987 and June 1988, 335 femoropopliteal bypass (FPBP) operations were performed on 308 patients who had limb ischemia caused by femoropopliteal occlusion. The greater saphenous vein graft (SVG) was used whenever possible (165) and polytetrafluoroethylene (PTFE) prosthesis was used in 165. In addition, 4 Dacron grafts and 1 umbilical vein graft were used in 5 operations but were excluded from comparison in this study. The two groups did not differ significantly regarding stage of peripheral ischemia, preoperative indication, distal popliteal anastomotic site, and number of patent runoff vessels. Overall cumulative patency rates according to occlusive criteria alone were calculated by the life table method. Graft occlusion was determined by angiography, Doppler assessment, loss of previously palpable pulses, or return of symptoms. Diabetes mellitus was associated with a significantly lower patency rate.