SUMMARYA group of 10 patients with complete occlusion of one internal carotid artery (COG) were compared with a group of 12 patients with cerebral infarction and severe atherosclerosis but no complete occlusion of an internal carotid artery (IOG). The two groups were closely matched clinically and by age, degree of atherosclerosis, time from infarction, cerebral blood flow, and metabolic studies while breathing room air. After breathing 5% carbon dioxide, the COG had significantly less increase in CBF and decrease in CVR. CMRO2 decreased in the completely occluded group and increased in the incompletely occluded group. It is postulated that in the area normally supplied by the occluded internal carotid artery there is a large area that functions because the arterioles are maximally dilated. Under normal conditions, cerebral blood flow is maximal but may be decreased to the point where oxygen supply is at a critical level. When 5% CO2 is inhaled, the increased PaCO2 directly affects the large amount of brain not involved and dilates the arterioles, so that blood that would normally be shunted into the area of maximum need is now stolen from the collaterals. As this area of brain is already receiving the maximum amount of oxygen that it needs for normal metabolism, the CMRO2 in this region does not change. In the area of maximum need, however, the flow has decreased far below the critical level, and as oxygen is not available for normal metabolism, the CMRO2 decreases in this area. Thus, the average cerebral blood flow may not be affected, but the average CMRO2 will be significantly reduced. It would seem that vasodilators may be contraindicated in most cases of complete internal carotid artery occlusion and that probably before they are used in any patient CBF, CMRO2, and CVR responses to 5% carbon dioxide should be determined.