A retrospective study was undertaken to see if screening ultrasounds and elective surgical revision could decrease access failure. Two hundred fifty-three accesses [177 gortex grafts, 76 arteriovenous (AV) fistulas] were studied with duplex imaging. Patients were subdivided by access type, flow, percent stenosis, and whether they were surgically revised. Data was examined to determine access failure within 6 months. Ten of 76 fistulas clotted (13.1%), while 53 of 177 grafts clotted (29.9%) (p = 0.005). In Gortex grafts, stenosis (p < 0.05) and decreased flow (p = 0.005) correlated with clotting. In unrevised grafts with flow < or = 801 ml/min, 13 of 14 (92.8%) clotted, whereas of those with flow > or = 1603 ml/min only, 10 of 38 (26.3%) clotted (chi-square = 24.74; p < 0.0001). Only 1 of 18 (5.6%) revised grafts with flow < or = 1300 ml/min clotted, while 29 of 69 (42%) unrevised grafts clotted (p = 0.004). We were unable to demonstrate decreased clotting in fistulae with revisions. In conclusion, screening duplex scanning was able to select groups with a higher risk of access failure over the subsequent 6 months. Elective revision with correction of areas of stenosis in grafts with flows < or = 1300 ml/min significantly decreased the incidence of clotting.