This study examines the mediational role of gender in the effects of two patterns of cognitive and self-regulatory strategy interventions in the writing self-efficacy calibration of students with learning disabilities (LD). 121 5th and 6th Primary grade students with LD (43 girls and 78 boys), ranging in age from 10 to 12 years old were randomly allocated either to one of the experimental intervention groups, (n = 48, 19 girls and 29 boys), and followed a intervention program based on the Self-Regulated Strategy Development Model, or they received training based on the Social Cognitive Model of Sequential Skill Acquisition (n = 31, 15 girls and 26 boys), or alternatively they were allocated to the ordinary instruction group (n = 32, 9 girls and 23 boys). Writing performance was assessed using two types of writing evaluation: a reader-based score concerned with structure, coherence and quality, and a text based score regarding productivity, coherence and structure. Writing self-efficacy beliefs were also assessed using a self-report scale including eight items about the students' confidence in completing a writing task and to gain specific writing skills. The results suggest that the miscalibration of writing self-efficacy in girls with LD was significantly modified to a more realistic calibration of their writing competence after experimental intervention. However, the findings do not confirm the same clear statement for boys.