Observational methods such as OWAS, RULA, and REBA have been widely used to identify posture-related risks of musculoskeletal disorders in industry, since they are useful and efficient in evaluating postural stresses. However. there are few studies comparing the methods and providing guidelines for selecting and using the methods. They have been developed based on different backgrounds and with different application areas. Each method has its own characteristics. which must be considered in selecting and using them. In this study. 17 male subjects evaluated 42 different working postures that frequently assumed in the automobile assembly line using a psychophysical method. The postures were then evaluated by different observational methods. The results of the observational methods were compared with psychophysically evaluated stresses. The observational methods resulted in different values of stresses for certain postures. For some postures showing high values of perceived discomfort. the observational methods showed different values of stresses. These results showed that the observational methods should be used differently according to application area and they have some weak points to be improved.