Our prior anecdotal experience with unsuspected preexisting saphenous vein disease prompted us to study its incidence, its relation to graft failure, and to identify techniques for its detection. Thick-walled, postphlebitic sclerotic occluded, postphlebitic sclerotic recanalized, calcified, and varicose vein lesions were detected in 63 (12%) of 513 infrainguinal vein bypasses. In 13 (2% to 5%) cases, severe saphenous vein disease precluded use of the vein. In the remaining 50 cases, the entire vein or a portion thereof, with minimal or unsuspected disease, was used for bypass. Early graft failures occurred in 10 (20%) of the 50 cases. The cumulative primary patency rate at 30 months for bypasses performed with diseased veins was 32%. This was significantly less than the 73% cumulative primary patency rate for bypasses with veins without detectable disease (p less than or equal to 0.001). Retrospective evaluation of preoperative duplex ultrasonography (n = 21) originally used to evaluate saphenous vein length and diameter correctly identified thick-walled, occluded, calcified, and varicose veins in 62% of cases. Intraoperative methods of vein evaluation included inspection, palpation, irrigation, catheter or valvulotome insertion to identify obstruction, and intraoperative arteriography. Histologic examination of diseased veins demonstrated a spectrum of disease with thickening of the intima and media, vein wall calcification, and luminal recanalization. We conclude that (1) unsuspected preexisting saphenous vein disease occurs in approximately 12% of cases and results in both early and late graft failures; (2) detection, in some cases, is possible with duplex ultrasonography and intraoperative techniques; and (3) diseased veins that are recanalized, calcified, or thick-walled should not be used if an alternative vein is available.