Experiments were conducted to clarify the question of endogenous ethanol. The sensitivity and specificity of gas-liquid chromatography has been exploited; with the use of the proper chromatographic column ethanol can be identified and its presence measured in vapor samples, such as alveolar air, at concentrations equivalent to 0.5 mg of ethanol per liter of blood. A column with Halcomid M- 108L as the active phase and a hydrogen flame ioniza-tion detector were used. Using alveolar air for sampling, no differences in ethanol concentration were noted between 19 nonalcoholics and 6 alcoholics. In 10 subjects there was no detectable alcohol; in 7, the concentration was equivalent to 0.5 mg/1 of blood; in 5, 1.0 mg/1; and in 3, 1.5 mg/1. The first direct determinations in human subjects of the rate of disappearance of ethanol at concentrations below 100 mg/1 of serum are reported. The rate of disappearance is linear until this level, or in some subjects as low as 50 mg/1, is reached. The rate of disappearance, in milligrams per liter per hour, may be calculated as the product of the concentration, in milligrams per liter, and the constant 1.64, the mean of 5 experiments in 3 subjects. Exposure of 6 human subjects to simulated altitudes of 4000 m for 1 to 2 hours and 4 rats to 7000 m for 5 hours did not produce any increase in the alcohol concentration or in any other carbonaceous material appearing in the expired air. Direct determination of the concentration of ethanol in rat liver gave a value of 9 mg/1000 g in contradistinction to 1/10 these values when peritoneally equilibrated nitrogen was used in the living animal. It is suggested that ethanol formation may take place during handling of tissues and thus account for the discrepancy. It is concluded that ethyl alcohol may be present in humans in concentrations up to 1.5 mg/1 of blood but whether this alcohol is of endogenous origin is unresolved; even if such formation were endogenous, its fraction of the basal metabolic rate would not exceed 1%. A possible role of endogenous ethanol in a theory of the etiology of alcoholism appears excluded.