This study investigates the effects of two aspects of writing revisions (revision amount and revision function) on writing improvement, and how students' self-rating accuracy moderates these effects. Hierarchical linear regression is employed with a sample of 114 undergraduate students. When considering the moderating role of self-rating accuracy, the results show that both revision amount and revision function had positive impacts on writing improvement; self-rating accuracy had different moderating effects. Revision amount enhanced the writing improvement of students with low self-rating accuracy, while revision function boosted improvement among students with high self-rating accuracy. The findings extend our knowledge of how writing revisions work alongside self-rating accuracy to promote writing. Pedagogical implications for tailored assistance provided to students with varying levels of self-rating accuracy are discussed.