Objective To analyze the situation and risk factors of non-suicidal self-injury（ NSSI） behavior in adolescents with nightmare disorder. Methods This study was a cross-sectional study. A questionnaire survey was conducted among students enrolled in five high schools and five middle schools in Ju County， Shandong Province， from May 1 to 30， 2022. The questionnaires were distributed via the Wenjuanxing platform. A total of 7 734 questionnaires were distributed， and 7 002 valid questionnaires were recovered， with an effective recovery rate of 90.54%. A total of 6 700 cases that met the inclusion criteria were selected. Based on scores of the Chinese version of the Nightmare Disorder Index（ NDI-CV）， participants whose answer to Item 1 was "0" were classified as the non-nightmare disorder group（ n=3 674）； those whose answer to Item 1 was not "0" but answers to Items 2-4 were "0" or "1" were classified as the subthreshold or partial nightmare disorder symptom group（ n=2 812）； those whose answer to Item 1 was not "0" but whose answers to Items 2-4 were "2" or higher were classified as the possible nightmare disorder group（ n=214）. The general information and Adolescent Non-suicidal Self-injury Assessment Questionnaire（ ANSAQ） scores of adolescents with different nightmare disorders were compared respectively， and post-hoc pairwise comparisons were conducted. The incidence of NSSI behavior in adolescents with possible nightmare disorder was recorded and divided into the NSSI group（ n=125） and the non-NSSI group（ n=89）. The NDI-CV， Insomnia Severity Index（ ISI）， Chinese Adolescent Daytime Sleepiness Scale（ CADSS）， Patient Health Questionnaire-4 items（ PHQ-4）， Self- Hate Scale（ SHS）， and Adverse Childhood Experiences Questionnaire-Revised（ ACEQ-R） were used to record adolescents' scores on sleep problems， emotional status， self-hatred， and adverse childhood experiences. Multivariate Logistic regression was used to analyze the influencing factors of NSSI in adolescents， and the risk prediction model was constructed. The receiver operating characteristic（ ROC） curve was drawn to validate the predictive performance of the model. Results The ANSAQ total scores for the non-nightmare disorder group， the subthreshold or partial nightmare disorder symptoms group， and the possible nightmare disorder group were 0（0，1.00）， 0（0，1.00）， and 6.50（0，17.00）， respectively， there were statistically significant differences in ANSAQ scores among adolescents in different nightmare disorder groups（ H=566.380，P ＜ 0.001）， and all post-hoc pairwise comparisons were statistically significant（ all P＜ 0.001）. One-way analysis revealed statistically significant differences between the NSSI group and the non-NSSI group in terms of scores for ISI， PHQ-4， SHS，ACEQ-R （Z=-2.174，-6.046，-12.671，-6.312； all P ＜ 0.05）. Multivariate Logistic regression analysis showed that anxiety and depression［ OR=1.170， 95%CI（ 1.011，1.354），P=0.035］， selfhatred ［ OR=1.084， 95%CI（ 1.047， 1.123），P＜ 0.001］， adverse childhood experiences［ OR=1.232， 95%CI （1.053，1.441），P=0.009］ were all independent risk factors for NSSI behavior among adolescents with nightmare disorder， and these differences were statistically significant. A risk prediction model for NSSI behavior among adolescents with nightmare disorder was constructed according to multivariate Logistic regression analysis. ROC curve analysis showed that the area under the curve for the risk prediction model was 0.838［ 95%CI（ 0.783， 0.894）， P ＜ 0.05］， with the sensitivity of 72.0% and the specificity of 85.4%， indicating good predictive performance. Conclusions The incidence of NSSI behavior in adolescents with nightmare disorder is high， and is affected by multiple factors. Anxiety and depression， self-hatred， and adverse childhood experiences were risk factors. The prediction model based on risk factors has good predictive value for the occurrence of NSSI behavior.