Cumulative graft patency rates calculated using graft occlusion as the end point are the standard method of presenting results of bypass surgery for lower limb ischaemia. The problems of using graft occlusion as the end point are that this is not easily documented and it gives no indication of the condition of the patient's limb after the graft occludes. The date of amputation is a well defined end point and it means treatment has failed. It is used to calculate cumulative limb salvage rates. Using the two techniques to assess different risk factors (age, calf vessel run-off, diabetes, position of distal anastomosis and hypertension), it was found that the limb salvage rate was a better indicator of patient progress. Whereas graft patency rates for diabetics and non-diabetics were similar (chi 2 = 0.8, P greater than 0.1), diabetics had a higher amputation rate and the limb salvage rate was significantly worse (chi 2 = 5.0, P less than 0.05). Cumulative survival is rarely presented in vascular series but it could be used as an indicator of the general condition of patients being selected for bypass surgery. The cumulative survival of diabetics was 23 per cent (s.e.m. +/- 12 per cent) at four years, while for non-diabetics this was 55 per cent (s.e.m. +/- 15 per cent), (chi 2 = 10.6, P less than 0.001). Diabetic patients have such different limb salvage and survival rates compared with non-diabetic patients that their results should be presented separately. A better indication of patient progress following bypass surgery is obtained if limb salvage rates and survival rates are reported as well as graft patency rates.