To determine a treatment strategy based on the outcomes of various previous interventions for critical limb ischemia in arteriosclerosis obliterans (ASO). We examined outcomes of 292 ASO patients who had had critical limb ischemia between May 1995 and July 2009. Patients underwent the following procedures in 167 cases: aortofemoral bypass (n = 14), femorofemoral crossover bypass (n = 29), femoropopliteal bypass (n = 104) and femorotibial bypass (n = 40). Other procedures included bypass only (n = 147), bypass combined with thromboendarterectomy (n = 10), bypass combined with endovascular therapy (n = 6), bypass combined with lumbar sympathectomy (n = 2), endovascular therapy combined with thromboendarterectomy (n = 4), endovascular therapy (n = 19), lumbar sympathectomy (n = 6), conservative therapy (n = 65), and major amputation (n = 31). We also calculated P3 risk scores and measured transcutaneous oxygen pressure (tcPO2) and skin perfusion pressure (SPP) before and after therapy. The limb salvage rate was 87% at 2 years in the arterial reconstruction group. In the low-risk group (a P 3 risk score of 3), the 1-year amputation-free survival rate was 96%. In the medium-risk group (a P 3 risk score of 4-7), the 1-year amputation-free survival rate was 88%. In the high-risk group (a P 3 risk score of 8), the 1-year amputation-free survival rate was 66%. The hospital death rate in the arterial reconstruction group was 3.2%, all of whom were patients who underwent bypass. The survival rate at 5 years was 65% and 36% in the conservative therapy only group. Ulcers healed in 140 out of 144 patients. The 4 patients with unhealed infections had tcPO2 or SPP values of more than 30 mmHg after treatment. Major amputations were performed in 4 of 5 patients who had tcPO2 or SPP values from 20 to 30 mmHg after treatment. Major amputations were performed in all 6 patients who had tcPO2 or SPP values of less than 20 mmHg after treatment. In cases with tcPO2 or SPP values of more than 30 mmHg, an ulcer will probably heal, except in infected cases. We suggest that, if these values are less than 30 mmHg, complete revascularization should be performed. The P3 risk score was useful in predicting limb salvage in the current series. Hybrid therapy in bypass and endovascular therapy must be performed in cases where patients are in a generally poor condition. It is important to attempt amelioration in limb salvage and to control the operative mortality rate with sufficient perioperative control. (English Translation of Jpn J Vasc Surg 2011;20:905-911).