Industrial fatigue may be understood as an out-put of lowered physiological functions and mental activities resulting from work. Our studies on fatigue have made clear correlations between various kinds of fatigue tests. Reasonable approaches to the evaluation of fatigue may be provided by our new method called "TAF test", the principle of which comes from an idea that fatigue may be grasped by observing ability in target-aiming as a representation of mental activities and ability in keeping up the concentration of aiming as a representation of physiological functions. In the present stage the examinee aims at the target as a distance of two meters with the gun-barrel for one minute continuously and rests for ten seconds, while in sitting he must avoid keeping his body stable by grasping the table with his free hand or by putting his elbow on the table or by touching his face with the gun-barrel to make aiming easier. In this manner aiming is repeated until he has aimed for a total of three minutes. When the target center is accurately sighted, the level is designated as O and in proportion to the inaccuracy in sighting the numerical figure increases up to 10. From each of the three curves shown during the three minutes, level (L) and deviation (D) are determined. L indicates the height of the average level of these curves and D indicates the standard deviation from L. In this sense the function of concentration maintenance can be expressed by the numerical values of L and D, and variances in TAF values due to metal and physical stresses are analysed statistically. (Jap. J. of Hygiene, 18 (2), 241∼246, 1963) For this study, 17 healthy male bank employees were selected. The average year of age was 23.5 and that of experience was about 4. The experiment covered 6 days from December 26 to 31, since this period is included in the busiest season of the year in Japanese banking business. There was overtime work daily from about December 10 to the last day of the year. Especially in the period selected for the experiment work ended usually at 8 to 10 p.m. During this period the density of work was also high and the actual working hours on most of the days were 10 to 12 hours. Because of this condition, pre-work values on the first day cannot be considered as normal. Therefore, for the control the pre-work values in the normal business time were used. The tests administered were TAF test (TAF), measurement of flicker frequency (FF), aesthesiometry (Es), blocking test in continuous colour naming (Bl), Donnagio reaction (Do), and Ogawa's reaction (O). In addition to these objective tests, the subjective fatigue questionnaire was also given. The results obtained were as follows: In the pre-work/control ratio (changes in pre-work values compared to the control values) FF, Es and Bl failed to indicate any lowering tendency except on the first day. TAF showed a lowering throughout the 6 days. The average difference in TAF L for each individual showed a significant lowering on every day except on the 4th and 5th days. In regard to changes in the post-work/pre-work ratio (changes during the day) FF, Es and Bl showed no lowering of functions. According to TAF, almost no changes in ratio occurred during the first 4 days, while in the last 2 days the lowering was observed. The average difference in TAF L for each individual showed a lowering tendency on every day, especially significant in the last 2 days. In urine tests, both O and Do showed changes that were either normal or slight. From the facts mentioned above, it may be said that the employees were fatigued already before the first day of this study, and that only toward the end of the week did the daily additional fatigue by work manifest itself as an accumulation of fatigue. As for the appearance ratio of subjective fatigue symptoms, the pre-work values were higher than those of the control. Values were also higher after work than before work in regard to this period.