The effects of flumazenil, a benzodiazepine antagonist, on the rate of cerebrospinal fluid (CSF) formation (&OV0312;f), resistance to reabsorption of CSF (Ra) and the electroencephalogram (EEG) was determined in 12 dogs anesthetized with halothane (0.4%, end-expired) and nitrous oxide (66%, inspired) in oxygen. In six dogs the responses to flumazenil were measured during administration of midazolam (1.6 mg/kg followed by 1.25 mg &OV0312; kg−1&OV0312; h−1, intravenously) given along with inhalational anesthesia, whereas in the other six dogs the responses to flumazenil were measured during inhalational anesthesia without midazolam. &OV0312;fand Rawere determined using ventriculocisternal perfusion, and EEG activity was evaluated using aperiodic analysis. Flumazenil, 0.0025 and 0.16 mg/kg, was administered both when CSF pressure was normal and when CSF pressure was increased to 36–38 cmH2O by continuous infusion of mock CSF. Flumazenil produced no statistically significant change in &OV0312;f. Flumazenil did produce inconsistent and relatively small changes in Ra. Quantitative aperiodic analysis indicated changes in EEG activity only when the larger dose of flumazenil was given to dogs receiving midazolam. At normal CSF pressure the changes were consistent and were comprised of decreases in &thetas;, α, and total hemispheric power. At elevated CSF pressure the changes were less consistent. It is concluded that smaller doses of flumazenil (which cause no EEG changes with the present method of analysis) and larger doses of flumazenil (which reverse midazolam-induced increase of &thetas; and α activity) produce no change of &OV0312;fand no consistent change of Ra. Although flumazenil given in the presence of midazolam may increase Ra, thereby increasing CSF pressure and impairing contraction of CSF volume, this effect is not likely to be clinically important.