The current study was conducted to examine the impact of attribution retraining on changing attitudes of primary school education. Attribution retraining as the independent variable and change of attitude as the dependent variable were considered. This study consists of field trials with pre-test, post-test and control group. Research instruments were Children Attribution Retraining Styles Questionnaire (CASQ) provided by Seligman and the researcher translated it and then revised it to match with the culture of Iranian school students.To investigate the influence of attribution Retraining statistical sample of 54 students were selected by multistage random sampling methods with two experimental groups and a control group. Pre-test was taken from both groups. Then, the experimental group was taught attribution retraining; after completing the course, post- test was administered to both groups. The analysis of one-way covariance in MANCOVA text showed that attribution retraining, significantly increased student’s attitudes, subseales of positive-permanent event, positive-comprehensive event, positive-private event, and positive events. Also,attribution retraining significantly reduced subscales of negative-permanent event and negative-private event, but it didnot significantly reduce negative comprehensive event (p< 0.01). Attribution retraining would change the students' attitudesand the resulting change may also have a great impact on various aspects of students' lives.