Various non-invasive vascular studies have been reported to provide valuable data for selection of the optimum level of amputation in limbs in patients who have vascular disease. We evaluated three such methods: (1) measurement of the change in the transcutaneous PO2 after inhalation of oxygen; (2) determination, by the Doppler method, of segmental blood pressure; and (3) measurement of the temperature of the skin. The records of eighty patients (ninety amputations) were retrospectively reviewed for correlations between the results of the vascular studies and the outcome of the amputation. Measurement of transcutaneous PO2 was found to be the most accurate predictor of successful healing of an amputation; the other two measurements were less reliable. The values for transcutaneous PO2 both at rest and after inhalation of oxygen were significantly different (p less than 0.001) for the patients who had a healed amputation compared with those who had a failed amputation. Regardless of the initial value, if, after inhalation of oxygen, the transcutaneous PO2 reached ten millimeters of mercury or more, it predicted healing of the amputation stump with a sensitivity of 98 per cent. When the level of amputation was selected on the basis of clinical judgment at the time of operation, the sensitivity was only 90 per cent.