The results of 93 consecutive lumbar sympathectomies performed over a five year period upon 54 patients with diabetes and 39 without diabetes were compared in terms of subsequent amputation using life table analysis methods. Cumulative success rates for avoiding amputation were significantly lower in those with diabetes as compared with rates for those without diabetes at five years and at most shorter six month intervals. Although there are a number of possible explanations for this difference, an important contributing factor may be the prior spontaneous denervation of diabetic blood vessels. The results demonstrate that better predictive indexes for lumbar sympathectomy are needed.