It is said that the category-words of a category scale must be differenciated from each other in strength of their semantic meanings and be arranged in order of strength of their semantic meanings. If the category-words not differenciated from each other in strength of their semantic meanings are used to construct a category scale, inconsistency may occur between the category-word order of the given category scale and the category-word order of the semantic meanings understood by the subjects and their judgment process with the category scale may be confused. The purpose of this study is to investigate the order effects of category-words on the subjects' judgment process with category scales which include the category-words undifferentiated from each other in strength of the semantic meanings. From our experiment, we get the following results: (1) The category-word order of a category scale must fulfill the following conditions: A. Expressive validity: The category-word order of a category scale must possess an expressive consistency. a. Example of the category scale with the high expressive validity. _??_ b. Example of the category scale without the validity. _??_ B. Validity of strength of semantic meanings: The category-words of a category scale must be differenciated from each other in strength of th e semantic meanings. Dochiratomoienai (neither long nor short) and amarinagakunai (not so long) should not be included in the same category scale, because the significant difference in strength of their. semantic meanings is not found between the words -55% of the 327 students judged “dochiratomoienai (neither long nor short)” is longer thri “amarinagakunai (not so long)” in paired comparison (X2=3.12, P>0.05)-. C. Validity of category-word order: The category-words of a category scale must be arranged in order of strength of their semantic meanings. Judgment process with a category scale which does not fulfill these three conditions must be confused. (2) when a category scale which possesses the high expressive validity but lacks the validity of strength of the semantic meanings is given, the subjects have a strong tendency to judge according to the category-word order of the given category scale. (3) When the category-word order of a category scale possessing the high validity of strength in the semantic meanings is not identical with the category-word order of strength of the semantic meanings understood by the subjects, the subjects rearrange the category-word order of the category scale according to their semantic understanding. So the expressive validity and the validity of strength of the semantic meanings are more important factors th an the validity of the category-word order in order to construct a category scale.