Surveillance of in situ saphenous vein bypass grafts with duplex scanning detects graft abnormalities which may lead to graft thrombosis. Correction of these defects, while grafts are still patent, potentially improves overall graft patency. In this study we compared color flow and conventional duplex to determine whether color flow provided additional information not obtainable by conventional duplex examination. The primary patency rate (patency maintained without intervention) for all 51 cases was 76% (39/51). The secondary patency rate (patency maintained by identification and correction of graft defects before failure) was 88% (45/51). Duplex scanning reduced the graft failure rate by 50%. Color flow and conventional duplex examination provided the same information regarding incipient graft failure. In 20 patients monitored with both techniques, the same number of proximal (100%) and distal (90%) anastomoses were imaged. The same number of graft defects (three vein graft stenoses, one proximal femoral artery stenosis) were identified. Velocity data obtained using the two techniques (peak systolic velocity in an area of stenosis and the duplex velocity ratio) were not always the same, making calculation of percent stenosis from this data inaccurate. Color flow duplex is useful in monitoring graft patency, but provides no additional information over that provided by conventional scanning.