The classic workup of patients considered for carotid endarterectomy (CE) has included contrast arteriography to delineate the nature and extent of the arterial pathology. Noninvasive testing (NIT) consisting of duplex scanning plus sound spectrum analysis is an alternative method for accurately evaluating the carotid bifurcation. The accuracy of our laboratory in comparing NIT to contrast arteriography has been established by use of the guidelines suggested by the Intersocietal Commission for the Accreditation of Vascular Laboratories. Forty-two patients underwent CE based on NIT alone. Seventeen had hemispheric transient ischemic attacks or minor strokes, six had amaurosis fugax, three had nonspecific symptoms, and 16 were asymptomatic. NIT identified a severe stenosis that was limited to the carotid bifurcation; the internal carotid artery distal to the bulb was normal. The predicted stenotic atherosclerotic lesion was confirmed at operation in all patients. One patient had a high bifurcation that required an extension of the original incision. Of three undiagnosed kinked distal internal carotid arteries found at operation, only one required surgical correction. No transient ischemic attacks, strokes, or deaths occurred postoperatively. CE can be performed in selected patients based on NIT alone, obviating the mortality, morbidity, and cost of arteriography. This algorithm demands that the NIT is unequivocal and the accuracy of the testing laboratory is established and maintained according to published standards.