It is very worthwhile to evaluate the grade of drunkenness objectively, until now, no reliable method has been established for this purpose. A new performance test called“TAF test”was devised by Takakuwa in 1960, and the utility of this test has been proved in evaluating fatigue as well as testing occupation aptitude. (Ergonomics, Vol. 5, No. 1, 37-49, 1962)The present studies were made in order to ascertain whether the TAF test could grasp the state of drunkenness quantitatively or not. In addition, the alcohol content in blood was measured with Lester's method and that in expiratory gas was measured with testing tube. The questionnaire on subjective drunken feeling (classified from 1st to 5th degree) was also administered. The experiment consists of 2 series: in the first series 16 male students ranging 21∼23 years of age were given 300ml of SAKE (Japanese wine, 30 proof alcohol) before supper, and in the second series 15 male students ranging 21∼24 years of age were also given 540ml of SAKE. In both series TAF test was administered to each examinee in the morning (control value), before the drinking (pre-value), 10 and 50 minutes after the drinking (post-value). Measurements of the alcohol content both in blood and expiratory gas and the questionnaire were also administered together with the tests after the drinking. The results obtained were as follows:1. 1st seriesAccording to the changes in TAF L, the examinees were divided into 2 groups. In one group, the pre-values were almost the same as the control values, but each of them showed a apparent lowering after the drinking (group N). In the other group, the examinees of which were undoubtedly affected with considerably strenuous lecture before the experiment, the pre-values were significantly lower than the control values, but each of them showed a significant increase after the drinking (group S). This increase may be due to the fact that the stresses to the examinees of the latter group were dispelled with the pharmacological actions of alcohol. In this series no correlation was seen between the changes in TAF and the blood alcohol content.2. 2nd seriesTen minutes after drinking, the blood alcohol contents were more than 60mg/dl in most of the examinees and each of them showed further increase up to 75∼155mg/dl after 50 minutes. The result of the questionnaire revealed that the examinees felt mostly the 3rd degree of drunkenness, but no apparent relationship was observed between the alcohol content in blood and that in expiratory gas. Concerning average TAF values, L were 1.8 (control value), 2.0 (pre-value). 2.7 (10min. post-value) and 3.2 (50min. post-value), D were 0.62, 0.68, 0.72 and 0.83 respectively. Significant differences in L and D were observed between the pre-value and 10min. post-value, and also between the post-values of 10min. and 50min. In this series, a significant correlation was noticed between the lowering in TAF L and the blood alcohol content in the range above 50mg/dl. Especially in the extent above 90mg/dl, all examinees showed the lowering in TAF L larger than 0.5 which was seen only in the state of apparent fatigue.In conclusion, our studies demonstrate that TAF test can successfully evaluate the grade of drunkenness, and that the level of 90mg/dl in blood alcohol content may be the first threshold where the concentration ability is obviously disturbed.