The results of ten patients with severe, intractable symptoms of arterial disease receiving spinal cord stimulation are reported. Six out of ten patients showed clinical improvement. Three of five patients with severe rest pain obtained complete or very marked relief and one of two patients with moderate rest pain in the legs obtained complete relief. The mean claudication distance in the ten patients increased from 65 to 212 metres during epidural stimulation of the spinal cord. Exercise tolerance as measured on a bicycle ergometer increased by 61%. These changes were associated with small increases in cutaneous and muscle blood flow. In those patients who responded clinically, the improvements seen were maintained as long as spinal cord stimulation was continued. There was no clinical response to transcutaneous (placebo) stimulation and four patients did not respond in any way to spinal cord stimulation. The improvements seen are unlikely to be due to either the natural history of the disease or to a placebo effect. The effect is probably due to antidromic stimulation of the central processes of the first order sensory neurons. It is suggested on the basis of animal studies that this effect may be mediated by release of prostaglandins as well as indirectly via pain relief.