Recent data from the North American Symptomatic Carotid Endarterectomy Trial revealed a 14.3% perioperative risk of stroke or death with carotid endarterectomy contralateral to a carotid artery occlusion. Since last reporting on this topic in the mid-1980s, the authors have reviewed 180 patients with occlusion of one internal carotid artery (ICA) and who underwent endarterectomy of the stenotic contralateral ICA operated from 1965 to 1984 (group A) compared with 135 operated on from 1985 to 1991. The two groups were similar with respect to age, sex, incidence of coronary artery disease, hypertension, diabetes and history of smoking, but group B had a significantly increased incidence of patients who were neurologically symptom-free before surgery (21.5% versus 7.8%, P < 0.001). The combined perioperative stroke or death rate for patients in group B was significantly lower than for those in group A (0.7% versus 6.7%, P < 0.01). Comparison of the operative techniques showed more frequent placement of intra-arterial shunt (52.6% versus 29.4%, P < 0.001) and increased use of general anesthesia (20.0% versus 9.4%, P < 0.01) in patients of group B. Analysis of the etiology of the complications, however, showed that shunting alone could not account for the improved results. Lower incidences of postoperative thrombosis, embolization and intracerebral hemorrhage were equally important.(ABSTRACT TRUNCATED AT 250 WORDS)