Between 1964 and 1978 sixteen patients underwent a carotid subclavian by-pass in the Akademisch Ziekenhuis of Gent. Most of these patients were submitted to tests to evaluate the importance of the carotid steal effect. The patency of the vertebral artery was found to influence these tests. Two of the three patients with an occluded vertebral artery had cerebral symptoms when the peripheral resistance was lowered in the involved limb. No positive tests could be found in the twelve patients with a patient vertebral artery. These results indicate that the carotid subclavian artery by-pass performed in the presence of an occluded vertebral artery does not protect the patient from the carotid steal effect.