There were two main purposes for this study. First, to understand a clear definition of what concrete behaviors were associated with academic cheating behavior in elementary school students. Second, to compare the attitudes of the seriousness of academic cheating behavior among students of different variables in elementary school. The variables included grade, sex, academic scores, moral scores teachers valued, social economic situation and the level of parents´ education. 558 4-6 graders, 287 male and 271 female, were randomly drawn from nine elementary schools in Taipei City for this study. Research instruments used in this study were: Learning Behavior Scale designed to measure the individuals´ judgement of the severity of academic cheating behavior and Tables of Students´ Fundamental Records designed to collect the variables mentioned above. The data obtained were analyzed by frequency distribution, test of goodness offit, one-way analysis of variance and multiple serial correlation. The major findings were summarized as follows: Ⅰ.There were four types of academic cheating behavior. First, to gain answers on exam situation. Second, to offer answers on exam situation. Third, to rise scores by other skills out of the situations. Fourth, the inappropriate learning behaviors on ordinary learning situation or homework assignments. Ⅱ.There was slight difference between different grades. Generally, there was no significant difference in attitudes between grades. Ⅲ.There was significant difference on sex. Female students considered each of the thirty forms of academic dishonesty as more serious than male students. Ⅳ.There was positive correlation between the attitudes of academic cheating behavior and academic scores. The more the students possessed in academic scores, the more serious the students thought on academic cheating behavior. Ⅴ.There was positive relationship between the attitudes of academic cheating behavior and moral scores teachers valued. The more the students were valued as high grade, the more serious the students thought on academic cheating behavior. Ⅵ.The relationship between the attitudes of academic cheating behavior and social economic situation (SES) was not significant. That is, no matter what SES the students stood, the attitude about academic cheating behavior was concurrent. Ⅶ.The relationship between the attitudes of academic cheating behavior and the parents´ education level of students was not significant on grade 5 and 6, but there was significant relationship in father´s educational level on grade 4, not in mother´s. That is, the higher the father´s´ education level of students was, the less serious the attitudes of students were about academic cheating behavior on grade 4. At the end of the thesis, results were discussed, the implications of this study in both education and further research in this field were also offered.