We performed a retrospective review of a vascular surgery quality assurance database to evaluate the perioperative and long-term morbidity and mortality of above-knee amputations (AKA, n = 234) and below-knee amputations (BKA, n = 720) and to examine the effect of diabetes mellitus (DM) (181 of AKA and 606 of BKA patients). All patients in the database who had AKA or BKA from 1990 to May 2001 were included in the study. Perioperative 30-day cardiac morbidity and mortality and 3-yr and 10-yr mortality after AKA or BKA were assessed. The effect of DM on 30-day cardiac outcome was assessed by multivariate logistic regression and the effect on long-term survival was assessed by Cox regression analysis. The perioperative cardiac event rate (cardiac death or nonfatal myocardial infarction) was at least 6.8% after AKA and at most 3.6% after BKA. Median survival was significantly less after AKA (20 mo) than BKA (52 mo) (P < 0.001). DM was not a significant predictor of perioperative 30-day mortality (odds ratio, 0.76 [0.39-1.49]; P = 0.43) or 3-yr survival (Hazard ratio, 1.03 [0.86-1.24]; P = 0.72) but predicted 10-yr mortality (Hazard ratio, 1.34 [1.04-1.73]; P = 0.026). Significant predictors of the 30-day perioperative mortality were the site of amputation (odds ratio, 4.35 [2.56-7.14]; P < 0.001) and history of renal insufficiency (odds ratio, 2.15 [1.13-4.08]; P = 0.019). AKA should be triaged as a high-risk surgery while BKA is an intermediate-risk surgery. Long-term survival after AKA or BKA is poor, regardless of the presence of DM.