Since the beginning of this century, in Europe and America, a number of researches have been produced on the methodological problems of rating essays written to be evaluated on particular points of view. The most significant feature of essay tests is that the quality of responses should be judged subjectively by human raters. In spite of the educational merits of this testing method, it is hard to maintain the reliability of the tests, because various types of noises can easily be generated in scoring process. It seems that reseachers have given only little attention to these types of problems in our country. This article is trying to review some of the theoretical and empirical aspects of these researches which are considered beneficial to these methodological problems.