Brain capillary perfusion was evaluated in the different states of the wake-sleep cycle-quiet wakefulness (QW), quiet sleep (QS), and active sleep (AS)-in rats. The extent of the perfused capillary network was determined by intravascular distribution of a fluorescent marker. Evans blue (EB); it remained unchanged across the three behavioral conditions, QW, QS, and AS. The anatomical network was assessed by alkaline phosphatase (AP) endothelial staining, which is known to underestimate the number of existing capillaries. The resulting number of AP profiles were, therefore, significantly lower than the number of EB profiles, but the percentage of AP-stained capillaries that were perfused (96%) was also unchanged across the behavioral conditions. The results indicate that no capillary recruitment accompanies the wake-sleep cycle. Capillary surface area is a relevant factor in determining exchanges across the blood-brain barrier. In the absence of capillary recruitment (relative constancy of the surface area), the CBF changes during sleep should preferentially affect flow-limited with respect to diffusion-limited transport.