The extraction of 15O labeled water by the brain during a single capillary transit was studied in vivo in 20 adult rhesus monkeys by external detection of the time course of the tracer subsequent to the internal carotid injection of 0.2 ml of whole blood labeled with H215O. The data showed that labeled water does not freely equilibrate with the exchangeable water in the brain when the mean cerebral blood flow exceeds 30 ml/100 g min-1. At the normal cerebral blood flow in the rhesus monkey (50 ml/100 g min-1), only 90% of the H215O is extracted during a single capillary transit. In addition, cerebral blood flow was determined with H215O and 133Xe in these monkeys using residue detection and employing the central volume principle. The data supported the hypothesis that a diffusible tracer, H215O, need not be in complete equilibrium between the phases of a system for the application of the central volume principle to be valid. Finally, the brain capillary permeability surface area product was computed from these data; it was approximately 0.023 cm3/sec g-1.