Neuropathological studies have shown increased cerebral spaces in alcoholics, yet, the effect of ethanol on cerebrospinal fluid (CSF) production is not known. We investigated the effects of ethanol on CSF production measured by ventriculocisternal perfusion (VCP) technique, in two groups (n=10 in each) of anesthetized, paralyzed and mechanically ventilated dogs. In group I, which served as control, VCP was performed with normal mock CSF. Ethanol (150 mg/dl of mock CSF, approximately 33 mM) was added to VCP in group II. Beginning 60 min after the start of VCP, CSF production was measured every 15 min for the next 4 h. In group I, mean (±S.D.) value for CSF production was 51±10 μl/min initially and decreased significantly but slightly with time, to the lowest value of 44±11 μl/min at the end of the experiment. In group II, values for CSF production were 41±8, 41±8, 41±8, 43±6, 43±8, 42±6, 42±8, 38±6, 37±6, 36±5, 36±5, μl/min, respectively, from 15 to 165 min. These values were invariably significantly lower than their respective mean values in the control group. Furthermore, when ethanol was withdrawn at the trough of CSF production (at 165 min), production significantly increased by about 40%. We conclude that ethanol at a concentration of 150 mg/dl (far below lethal levels) is one of the most potent inhibitory drugs for decreasing CSF production. This effect is short-onset and is fully reversible within 15 min of ethanol withdrawal.