The purpose of this study was to establish whether carotid-revascularized patients who had preoperative vertebrobasilar insufficiency (VBI) displayed distinctive characteristics and whether a particular prognosis would ensure. From January 1985 to December 1993, 1022 carotid revascularizations were performed, of which 114 (11%) were for high-grade stenosis associated with VBI. The group with VBI and the group without VBI were compared according to a set of 121 prospectively collected variables. Of all the demographic and risk-factor variables, only female prevalence (42% vs. 27%) and hypertension (77% vs. 27%) distinguished the group with VBI, who also exhibited a significantly higher proportion of significant contralateral carotid lesions (27.2% vs. 8.9%) and vertebrosubclavian lesions (38.6% vs. 24.8%). Following isolated carotid surgery, there was no statistically significant difference between the two groups as to their cumulative rate of permanent neurological mortality and morbidity (2.6% in the group with VBI vs. 3.4% in the group without it). With an average follow-up of 60 months, VBI was cured in 82.4% and improved condition shown in 6.5% of patients. However, the proportion of good results fell to 65% in patients with a nonfunctional circle of Willis. Out of 13 cases of failure to control VBI, cure was finally effected by means of contralateral revascularization in 3 cases and by means of vertebrosubclavian revascularization in 5 cases out of 6. At 5 years, the actuarial rates of neurological event-free intervals and survival were not different from one group to another. In most cases, isolated carotid surgery is sufficient to bring vertebrobasilar insufficiency under control, except when significant vertebrosubclavian lesions and a nonpatent circle of Willis call for simultaneous carotid and vertebral artery surgery.