Purpose and Methods: The contribution of color-flow duplex surveillance to improving vein graft patency was evaluated in two patient groups after 201 infrainguinal bypass procedures were performed during a 5-year period. Incidence of revision procedures and the primary and assisted primary patency rates were compared for 160 bypass grafts monitored during the first 2 years by use of color-flow duplex scanning of the vein graft and adjacent arterial segments (color-flow surveillance group) versus 41 bypass grafts monitored by use of clinical assessment alone (clinical follow-up group). Only grafts that were patent after the first postoperative month are considered. Results: The two groups were comparable with regard to most of the pertinent clinical factors. Stenotic lesions were identified in 58 bypass grafts, and severity was determined by use of intraarterial digital subtraction angiography. Eighteen bypass grafts with stenoses did not undergo a revision for reasons that were determined by the doctor, the hospital, or the patient. The occlusion rates of revised and nonrevised stenotic grafts were compared for lesions of different severity. None of the grafts for stenoses with 30% to 49% diameter reduction (DR) failed during follow-up. Occlusion occurred in 57% of the nonrevised and 9% of revised grafts (p = 0.047) for stenoses with 50% to 69% DR. Stenoses with 70% or greater DR were associated with graft failure in 100% of nonrevised bypasses and in 10% of revised grafts (p = 0.004). The assisted primary patency rate was higher in grafts that underwent color-flow surveillance compared with grafts with that underwent clinical follow-up (3-year patency rates of 91% and 72%, respectively; p = 0.004). The independent correlation of color-flow surveillance with higher patency rates was demonstrated in a proportional hazard analysis. The relative risk (probability of occlusion) in color-flow surveillance grafts is less than one third of the relative risk in bypass grafts that underwent clinical follow-up. Conclusions: We conclude that revision procedures were more optimally used during color-flow surveillance, whereas asymptomatic stenotic graft lesions are missed with clinical follow-up, which results in a higher percentage of graft failures. Overall graft patency rates can be improved with use of color-flow duplex surveillance and repair of significant stenotic lesions. (J VASC SURG 1993;17:42-53.)