Objective: To explore the latent classes of peer victimization among junior high school students, and to analyze the association between the characteristics of different classes and non suicidal self injury(NSSI) behavior,so as to provide scientific basis for the prevention and intervention of NSSI behavior among junior high school students. Methods: A census method was used to select 19 738 students from 16 junior high schools in a central urban area of Wuhan as research subjects from November to December 2024. The Multidimensional Peer Victimization Scale and a self designed NSSI questionnaire were used to assess peer victimization and NSSI behaviors. Latent profile analysis (LPA) was used to identify latent classes of peer victimization among students by sex. Binary Logistic regression model was used to analyze the association between latent classes of peer victimization and NSSI behavior among junior high school students of different genders. Results: Latent patterns of peer victimization among boys were classified into 4 groups:no victimization group (85.8%, n =8 874), low victimization-verbal type group (6.6%, n =682), low victimization-mixed type group (5.7%, n =587) and high victimization group (1.9%, n =195); girls were classified into 4 groups:no victimization group (87.0%, n =8 179), low victimization group (7.5%, n =702), high victimization-social type group (3.8%, n =360) and high victimization-physical type group (1.7%, n =159). After controlling for only child status, place of residence, parenting style, and self rated family socioeconomic status, binary Logistic regression analysis showed that the risk of NSSI behavior among boys increased sequentially in the low victimization-verbal type group( OR =4.14, 95% CI =3.22-5.31), low victimization-mixed type group ( OR =4.75, 95% CI =3.68-6.13), and high victimization group ( OR =8.26, 95% CI =5.83-11.71) compared to the no victimization group; the risks of NSSI behavior among girls increased sequentially in the low victimization group ( OR =3.06, 95% CI = 2.51-3.72), high victimization-social type group ( OR =4.01, 95% CI =3.13-5.15), and high victimization-physical type group ( OR =6.18, 95% CI =4.38-8.73) compared to the no victimization group (all P <0.05). Conclusions There are gender differences in the latent classes of peer victimization among junior high school students and the risk of NSSI behavior varies across different victimization classes. Targeted prevention and intervention should be implemented based on the gender of junior high school students and the characteristics of peer victimization students experience.