To investigate the effect of interpersonal relationships on non-suicidal self-injury (NSSI) in adolescents with first episode depression. Methods: Eighty adolescent patients with first episode depression were enrolled and evaluated by the adolescent self-injury questionnaire for NSSI. According to the presence or absence of NSSI, patients were divided into NSSI group and no-NSSI group. All subjects were assessed with Hamilton Depression Scale (HAMD-17) and Hamilton Anxiety Scale (HAMA) for depression and anxiety symptoms. Parent-child intimacy scale, peer relationship scale, and teacher-student relationship scale were used to evaluate the interpersonal relationship of the samples, and an inter-group comparison was conducted. The independent risk factors of interpersonal relationships for NSSI were analyzed by binary Logistic regression. Pearson correlation analysis was used to analyze the relationship between NSSI and interpersonal relationships. Re? sults: 42 cases (52.5%) were enrolled NSSI group and 38 cases (47.5%) in no-NSSI group. Compared with the no-NSSI group, cases in the NSSI group scored higher in HAMD-17, HAMA, peer terror inferiority complex, and conflict of teacher-student relationship (all P<0.01), and lower in parent-child intimacy, father intimacy, mother intimacy, the total score of peer relationship, peer acceptance, intimacy of teacher-student relationship, support of teacher-student relationship and satisfaction of teacher-student relationship (all P<0.01 or P<0.05). Lo- gistic regression analysis showed that peer acceptance, peer fear and inferiority, father intimacy, and mother intimacy were independent risk factors of NSSI. Pearson correlation analysis showed a negative correlation between NSSI score and maternal intimacy in the NSSI group (P<0.01). HAMD-17 and HAMA scores were positively correlated with NSSI scores (P<0.05 or P<0.01), conflict between teachers and students, peer terror and inferiority complex (P<0.05 or P<0.01). Among the factors and indexes of HAMD-17, the anxiety/somatization factor was positively correlated with peer terror inferiority complex and NSSI score (all P<0.01); cognitive impairment factors were positively correlated with the conflict of the teacher-student relationship and NSSI score (all P< 0.05), and negatively correlated with the intimacy of teacher-student relationship, father intimacy and mother intimacy (all P<0.05); the retardation factor was positively correlated with the conflict of the teacher-student relationship and peer terror and inferiority complex (all P<0.05). Conclusion: Adolescents with depression have a high incidence of NSSI behaviors, which is related to interpersonal relationships. The poor teacher-student relationship, poor peer relationship, and the poor parent-child relationship are more likely to lead to NSSI behaviors.