The place of lumbar sympathectomy in the treatment of occlusive arterial disease of the lower limbs remains controversial. An adequate assessment of sympathetic nerve function in the practice of sympathectomy may reduce unsatisfactory results of this procedure. We report a series of 25 lumbar sympatholyses indicated for claudication, rest pain or distal arterial ulcer, and attempted by means of phenol injection in sympathetic ganglia under tomographic guidance. Sympathetic skin response (SSR) recording, an electrophysiological test of sympathetic activity, was performed before and after phenol injection, and its results were compared to the clinical outcome. SSR was abolished in the treated limb after phenol injection in only 64% of the cases. The post-injection abolition of SSR was correlated with the clinical improvement rate: 11% when SSR was still present, but 75% when SSR was abolished. In this latter case, the pre-injection values of SSR amplitudes could predict the clinical benefit of sympatholysis. It is necessary to assess the sympathetic activity in a limb before and after doing sympatholysis. Preoperatively, this activity may be diminished or already abolished in patients suffering from neuropathy involving the autonomic nervous system, limiting the indication of sympathectomy. Postoperatively, if sympathetic activity is still present in the treated limb, it implies that the sympathectomy is not completed, and it supports the indication of a new sympatholysis attempt. These observations show the usefulness of SSR recording in the practice of phenol sympatholysis, this test being simple and specific of sympathetic nerve function.