Non-suicidal self-injury (NSSI) poses a significant threat to adolescents with major depressive disorder (MDD) owing to elevated hospitalization and suicidality risks. Intermittent theta burst stimulation (iTBS) shows promise for treating NSSI in adolescents with depression. The efficacy of iTBS for treating NSSI in adolescents with MDD was evaluated in this study. In this sham-controlled, randomized clinical trial, adolescents with MDD were allocated randomly to active or sham treatment groups in a 1:1 ratio. They received active or sham iTBS over the left dorsolateral prefrontal cortex (LDLPFC) for five sessions daily for 5 days. The primary outcome was the difference in NSSI frequency and severity at 4-week follow-up, assessed using the Deliberate Self-Harm Inventory-Adolescent Revised version (DSHI-AR). Additional measures included changes in the Deliberate Self-Harm Ideation Scale for Adolescents Revised (DSHI-AR ideation) and Ottawa Self-Injury Inventory-addiction subscale (OSI-addiction). Of the 60 participants (mean [standard deviation] age, 14.2 [1.5] years; female sex, 55 patients [91.7%]), 85% completed the intervention. At 4 weeks post-intervention, the active treatment group showed greater DSHI-AR score reduction than the sham treatment group (mean difference, -18.66; 95% confidence interval [CI], -28.35--8.97). Active treatment led to a greater decline in DSHI-AR ideation (mean difference, -9.23; 95% CI, -14.41--4.05) and OSI-addiction (mean difference, -6.16; 95% CI, -10.58--1.74) scores compared with the sham treatment. Mild headaches were reported during the intervention, without significant group differences. The findings indicate that iTBS targeting the LDLPFC is an effective and well-tolerated strategy for treating NSSI in adolescents with MDD.