Remodeling occurs in diseased human coronary arteries; however, reports of remodeling in diseased autologous saphenous vein bypass graft (SVG) stenoses are inconsistent. Preintervention intravascular ultrasound and quantitative coronary angiography were used to study 104 SVG stenoses in 93 consecutive patients. Lesion site and proximal and distal reference segment measurements included vein graft, external elastic membrane, lumen, wall (vein graft minus lumen), and plaque (external elastic membrane minus lumen) areas. Three indexes of remodeling were assessed: (1) lesion site SVG (or external elastic membrane) area was compared with the average reference segment, (2) SVG area was correlated with the wall area and external elastic membrane area was correlated with the plaque area, and (3) the impact of excess plaque accumulation (at the stenosis compared with the reference segment) on lumen compromise was calculated. Overall, the ratio of lesion/reference vein graft area was 1.07 +/- 0.25; however, 23 lesions were classified as negative remodeling (ratio <0.9), 37 as intermediate remodeling (ratio between 0.9 and 1.1), and 44 as positive remodeling (ratio >1.1). Reference segment vein graft area correlated with wall area (r = 0.906, p <0.0001), and external elastic membrane area correlated with plaque area (r = 0.703, p <0.0001). Similarly, lesion site vein graft area correlated with wall area (r = 0.978, p <0.0001), and external elastic membrane area correlated with plaque area (r = 0.961, p <0.0001). The regression line relating delta lumen area to delta wall area was y = -0.22 x - 6.2 (r = 0.451, p <0.0001) and the regression line relating delta lumen to delta plaque area was y = -0.47 x - 4.5 (r = 0.572, p <0.0001). (A slope of 0 would indicate perfect positive remodeling and a slope of 1.0 no positive remodeling.) Diseased SVGs undergo positive and negative remodeling similar to native coronary arteries.