由結果得知：(1)舒適度門檻值將隨不同振動頻率而改變(0~1Hz)；(2)處於前後振動方向較處於左右振動方向敏感。然而，依據本論文之方法求得之回歸週期並無法合理地隨反對百分比增加而增長。此乃因為假設包松分配並不隨振動情況之不同而變，未來之研究將對此做適當修正，使回歸週期之決定更合理。 Under wind disturbance, the comfort of occupants in buildings is an important issue in the structural design aside from safety consideration. The human comfort criterion for wind-induced motion has been a long-time controversial issue because many factors, either physiological or psychological, can considerably differentiate the outcome. Recently, a constant criterion of 5 cm/sec2 for a half-year return period was regulated in Taiwan building wind code. In an attempt to build up the local data base in preparation for improving the building wind code in the future, this research aims to conduct surveys to investigate the comfort criteria by using a motion simulator newly constructed at the wind engineering research center, Tamkang University. A methodology and feasible procedure to determine the comfort threshold and the return period were proposed in this study. Two surveys (one tentative and one formal) using harmonic motion have been performed to obtain the comfort thresholds for 2%, 10% and 50% of people objecting and their corresponding return periods. The results show that (1) comfort thresholds are frequency-dependent within 0~1 Hz; (2) people with body orientation in fore-aft direction are more senitive to vibration than in side-to-side direction. However, it is found that, following the methodology proposed, the return period obtained does not reasonably increase with percentage of people objecting. It is due to the assumption of the same Poisson distribution for each vibration case in the methodology, which is to be improved in the future consecutive studies.