BACKGROUND Although autogenous arteriovenous fistulae are the optimal route for dialysis access, extended polytetrafluoroethylene (ePTFE) grafts continue to be the preferred access for patients without suitable superficial veins. Among the common complications related to dialysis grafts, thrombosis due to graft outlet stenosis is the most frequently encountered clinical problem. A cuffed graft was designed to eliminate the outflow turbulence to reduce outlet stenosis and to enhance the clinical patency of ePTFE grafts. We conducted a prospective, randomized study to compare the clinical outcomes of cuffed ePTFE grafts and noncuffed grafts in dialysis access. METHODS Between November 2004 and October 2005, 89 ePTFE grafts were implanted for hemodialysis access in the upper extremities of end-stage renal patients. Graft selection was randomized, with patients receiving a cuffed graft (Venaflo) or a regular noncuffed graft (Stretch Gore-Tex). All patients were monitored for signs of thrombosis or other complications. Primary and secondary graft patency was analyzed by using a life-table analysis, and the log-rank test was applied to compare graft patencies. RESULTS Demographic data for both groups were similar without statistical difference. The primary patency rates and secondary patency rates at 12 months after implantation were 56% and 91% for cuffed grafts, and 41% and 78% for noncuffed grafts, respectively. The cuffed group outperforms the noncuffed group regarding primary and secondary patencies statistically. However, the incidence of other complications that required further surgery was similar in both groups. CONCLUSIONS This investigation revealed that the cuffed ePTFE graft, which was designed to decrease graft outlet stenosis, may enhance the clinical patency rates of dialysis grafts.