Dose-response effects of acute ethanol infusions were studie, noninvaseively, in the unopened brain to examine the hypothesis that ethanol can induce stroke-like events as a consequence of cerebral vasospasm and tissue ischemia. By using a single sending and receiving fiber, and optical backscatter meauserment (500-800 nm) was used to monitor the levels of deoxyhemoglobin (DH), reduced cytochrome oxidase (rCO), and relative tissue blood content in a closed cranium preparation. Anesthetized rats were prepared by cannulating a branch of the internal cartoid artery and subjected to either bolus infusions (1.25 or 2.5 ÁM/g/min). To facilitate optical penetration, a portion of the left parietal cranium was shaved to a translucent appearance. Results showed that low, bolus doses of ethanol typically produced a slight increase (5-10%) in the oxyhemoglobin signal, indicating that vasodilation had probably occured. Higher doses, however, produced a promt and significant reduction in the hemoglobin signal, increased levels of DH, and a rise in rCO suugesting a vasoconstrictor response leading to ishchemia had occured, followed by recovery within 3-5 min. Constant infusions of ethanol produced a similar cerebral vascular response, in a dose-related manner, but of a more sustained nature. At levels of 50-60% of the maximum bolus dose, the effect was more pronounced, accompanied by an increase in the levels of rCO (by 50-90%). Control experiments using identical volumes/flow rates of Ringers solution produced solution produced no significant alterations in the optical spectrum. Overall, these data indicate that, depending on dose: (a)ethanol can induce vasodialatory or vasoconstrictor effects in the intact brain; (b) the more pronounced effects involve vasospasm in the cortical microcirculation leading to global ischemia; and (c) optical measurements permit direct noninvasive assessment of cerebral vascular effectts of alcohol and, potentially, other substances of abuse.