Serial arteriograms were obtained in 501 patients after coronary bypass grafting. Study I within 5 years of operation (mean interval 15 months) and Study II more than 5 years after (mean interval 88 months, range 60 to 147 months). One hundred patients received both internal mammary artery and saphenous vein grafts: 37, mammary artery grafts only, and 364, vein grafts only. In Study I, 645 (82%) of 786 vein grafts were patent, 42 (5%) stenotic or irregular, and 99 (13%) occluded. Of 140 mammary artery grafts, 136 (97%) were patent, two (2%) stenotic, and two (2%) occluded. Of the 645 vein grafts patent in Study I, 357 (55%) remained patent in Study II, 119 (18%) were stenotic or irregular, and 169 (26%) were occluded. Of 136 mammary artery grafts patent in Study I, 130 (96%) were unchanged, one was stenotic, and five (4%) were occluded in Study II. Early vein graft patency was influenced by the coronary artery grafted and by angina. Progression of vein grafts patent at Study I to stenosis or occlusion at Study II was associated with increasing postoperative interval (p less than 0.00001), interval myocardial infarction (p less than 0.001), angina (p less than 0.001), diabetes (p less than 0.004), hypercholesterolemia (p less than 0.006), and hypertriglyceridemia (p less than 0.02); it was not influenced by the coronary artery grafted. Within 5 years of operation, mammary artery graft patency exceeded vein graft patency. Between 5 and 12 years after operation, the attrition rate of vein grafts greatly exceeded that of mammary artery grafts (p less than 0.0001).