Investigation with actual trials about the bundle payment system is currently under way in Japan. The disease classification used for the investigation still plays an important role as a basis for analyzing the investigated results since the classification is closely related with other elements. Since it was our natural conclusion that failure of the disease classification to properly function made the proper preparation and operation of the intended system difficult, we investigated by re-analyzing information obtained in the DRG-related investigations and studies conducted in the past in Japan whether the disease classification used for the investigation met the intended purposes to make it suitable for the investigation, or what should be changed if the classification was not suitable. We focused our analyses on the number of days spent in hospital and their dispersion instead of the consumption quantity of resources, and confirmed that the classification in 183 categories being used for the investigation was rather small in its variation coefficient, and not inferior in comparison with AP-DRG of the United States of America and other Japanese disease classification systems. But since only a classification system having a small variation coefficient was deliberately selected to be used for the investigation, the above was the inevitable conclusion even taking into consideration the fact that no resource consumption quantity information was included in developing the disease classification system. It is possible that the good parts of the current classification system may be lost in the process of revisions, and therefore, it is suggested that careful study of how to revise the classification system and its purpose is needed.