Conflicts are inevitable in our daily life and should be solved in a constructive and creative way. In this study, mediation was identified as the way to resolve conflicts in a constructive way and has given some examples from the schools to show the implementation of "the model of mediation." The other objective of the study is to prepare a training program in which the mediation skills were taught to 7th grade students, and to research the effects of the program on the students' self concepts, leadership skills, aggressive behaviors, and the number of problems they perceived. It is expected that the training program would have positive effects on the students' self concepts and leadership skills, and would decrease the aggressive behaviors and the number of problems they perceived. CPI Do scale was used to measure the students' leadership skills, Piers-Harris Children's Self -Concept Scale was used to determine the level of students' self -concepts, Sears Aggression Scales were used to "identify the students' aggressive behaviors, and Mooney Problem Check List was used to determine the number of problems they perceived. According to the students' grades, the results of PARI, and the level of parents' socio-economic and cultural status, 25 students of 7th grades were selected as an experimental group, and another group of 25 were also selected as a control group. A 15 hours training program in which the mediation skills were taught was given to the experimental group. The training program consisted of the concept of coflict and different resolving methods, active listening, verbal andnonverbal communication, I -messages, the reflecting of feelings and understanding of the others' feelings, empathy, and the mediation process. The program was given to the 7th grade students in the periods of one and a half hours a week and lasted 10 weeks. The tests mentioned above were given to both of the groups before and after the training program. One-way variance analysis was used to determine the cumulative difference between the pre and post tests and, t-test was used to determine the differences between the groups. The statistical results showed no meaningful differences between the pre-tests and post-tests, and no differences between the groups. However the number of problems the students preceived increased in the control group, no increase occured in the experimental group.