This study aimed at investigating the incorporation of self-assessment in writing and its influence on improving writing skill. Forty-one Iranian female EFL learners participated in the study. Before starting the treatment learners had no idea of assessing their own writings and the task of assessment was done by the teacher like most of traditional classes. At the beginning of the treatment learners were asked to write a piece of writing and assess it. They were not instructed how to do that. The criteria for assessing writing and modeling it were practiced in the following session. For four successive weeks four pieces of writing were written and given marks by the learners based on the discussed criteria and the teacher corrected their writings and gave her own marks. The data were collected through self-assessment questionnaire administered to them before and after the treatment and learners' and the teacher's marks on writings. The data were analyzed by paired t-test in order to compare the test results before and after the treatment. A repeated measure of ANOVA compared students and the teacher's marks. The results of the analyses of the repeated measure showed students' writing skill improved gradually in the treatment period. The interviews with some of the learners and the teacher confirmed that learners and the teacher had positive attitude towards self-assessment.