The use of prosthetic grafts as vascular access for chronic hemodialysis is frequently necessary in patients with end-stage renal disease. Most commonly, expanded polytetrafluoroethylene (e-PTFE) has been employed because of ease of handling, tissue inertness, and acceptable long-term patency. Delay in use to allow for tissue ingrowth, however, has often required placement of temporary access devices. The authors have undertaken evaluation of a new material, plasma polymerized woven dacron Plasma-TFE, in a prospective randomized trial (Plasma-TFE VA) to compare clinical behavior against e-PTFE grafts, and we have used the Plasma-TFE grafts in an additional group of patients (Plasma-TFE AVA) as early access (within 1 week of implantation). Twenty-one Plasma-TFE grafts were implanted in 19 patients and 19 e-PTFE grafts were implanted in 17 patients in a prospective randomized fashion. Additionally, 31 Plasma-TFE grafts were implanted in 31 nonrandomized patients for early access. Primary patency rates in Plasma-TFE VA and e-PTFE grafts were equivalent at 12 months (0.471 and 0.556). When Plasma-TFE AVA primary patency was included (0.621), comparisons were not statistically significant (p = 0.50). Similarly, secondary patency rates among the three groups did not differ (cumulative proportion patent at 12 months: Plasma-TFE VA 0.403, e-PTFE 0.658, Plasma-TFE AVA 0.510). In considering after-revision patency after graft thrombosis, however, the Plasma-TFE grafts (both VA and AVA) performed significantly more poorly (p = 0.027) than e-PTFE grafts. Incidence of graft infection, wound infection, arm edema, hematoma from use, and occurrence of distal limb ischemia between Plasma-TFE (VA and AVA) and e-PTFE did not differ statistically. The authors conclude that Plasma-TFE compares favorably to e-PTFE with respect to primary and secondary patency and nonthrombotic complications, even with early use. Plasma-TFE does not perform as well as e-PTFE, however, after graft thrombosis.