The product of the permeability x vascular surface area (PA) of the blood-brain barrier to [14C]sucrose has been measured in rats maintained for 3 weeks in a chamber, the air supply to which carried a controlled concentration of ethanol vapour. No statistically significant difference was found between the permeability measurements in rats inhaling ethanol vapour for 3 weeks and non-alcohol exposed rats. The PA value was found to be significantly increased (115%) in rats given the same ethanol exposure when additionally subject to starvation during the last 3 days of this treatment. If the ethanol supply was also withdrawn at the same time as the food, a similar significant increase (116%) in PA value was found. In the absence of any ethanol exposure, 3 days' starvation did not significantly alter the measured PA value. Finally, when rats are given 200 mg/kg disulfiram every second day during a 2-week period of ethanol inhalation, the PA value was not significantly altered, although the concentration of acetaldehyde in the blood was up to 129 microM. The results indicate that while ethanol or acetaldehyde alone do not cause a weakening in the blood-brain barrier, the additional stress of food withdrawal after alcohol exposure does reduce barrier function, and this could be significant in human binge drinking.