OBJECTIVE There are excellent arguments in favor of the preferential use of prosthetic grafts above the knee for the treatment of infrainguinal occlusive disease. This approach has been popularized on the basis of the seemingly acceptable results when using polytetrafluoroethylene (PTFE). However, in many centers, knitted Dacron polyester has been used in these patients, and there are several studies that show equivalent and, in some, superior results with Dacron when compared with PTFE. The purpose of this study was to examine these results in a definitive way. METHODS A randomized prospective trial in eight clinical academic centers in the United States and Canada was initiated in 1991. Two hundred forty-four patients eligible for such a study, by virtue of criteria extant in each institution at the time, were centrally randomized. They underwent placement of either a knitted Dacron polyester graft impregnated with collagen or a thin-wall expanded reenforced PTFE graft to the above-knee popliteal artery, usually from the common femoral artery. They were frequently observed by protocol for as long as 5 years by a physical examination noninvasive hemodynamic study, including duplex scanning in many instances. Continuing patency was noted, as were other potential adverse outcome events. The data were analyzed by the log-rank test for cumulative patency and expressed as Kaplan-Meier curves. Data were further analyzed with a Cox proportional hazards model. RESULTS There were no differences in graft groups in demographic or comorbid factors. The procedural mortality rate was zero, and the morbidity rate was low (6.5%). The long-term patient survival rate was excellent (77% at 3 years). At the end of these years, no statistical significance in primary or secondary patency rates was observed between the two grafts (primary patency rate, 62% +/- 14.4% for Dacron; 57% +/- 15.5% for PTFE). No unexpected adverse outcomes on limb status were noted. Patency rates in both graft groups were inferior in patients who received small grafts (5 to 6 mm vs 7 to 8 mm; hazards ratio, 4.15) and younger (<65 years) smoking patients. CONCLUSIONS The fact that these two prosthetic grafts performed in equivalent fashion in a controlled, well-conducted prospective study is not surprising in spite of the previous work that suggested differences. If the preferential use of synthetic bypass grafts above the knee is to be used, it should be restricted to older nonsmokers with favorable anatomy. In that instance, the choice of graft material will depend on handling characteristics and cost. Above-knee prostheses should be only selectively used in younger, smoking patients, and graft size should be carefully considered in patients who undergo this operation.