Objective: To investigate the association between injury prevention knowledge level and injury occurrence among primary and secondary school students in Dongguan, so as to provide scientific basis for formulating targeted and evidence based intervention strategies for child injury. Methods: From March to April 2024, stratified cluster random sampling method was applied to select 5 988 primary and secondary school students in grades 4-9 for a survey. The Revised Questionnaire for the 2016-2020 Child Injury Prevention Project (senior version) was used to collect demographic characteristics, family and caregiver information, injury event characteristics, and knowledge, attitude, and practice related to injury prevention. Mann-Whitney U test was used for inter group comparison. Propensity score matching (PSM) was applied to balance confounders, and conditional Logistic regression was used to analyze the association between injury prevention knowledge score and injury occurrence among primary and secondary school students, and a restricted cubic spline (RCS) model was employed to test the dose response relationship. Furthermore, interaction terms were introduced to examine the moderating effects of sex, grade level, and caregiver’s educational level. Results: The overall incidence of injuries among primary and secondary school students was 18.15%. The knowledge score of injury prevention was significantly lower in the injury group [20.0(18.0,22.0)] than in the non injury group [21.0(19.0,22.0)] ( Z = 8.24 , P <0.01). After PSM, 1 068 pairs were successfully matched, and the standardized mean difference for all covariates was <0.1. An increase in injury prevention awareness scores was associated with a reduced risk of injury occurrence among primary and secondary school students ( OR =0.92, P <0.05). The RCS analysis showed an approximately linear negative association between injury prevention knowledge score and injury risk among primary and secondary school students ( P non linear =0.72), with a decreased injury risk when score exceeded 24 ( OR = 0.69, P <0.05). Subgroup analyses indicated that the negative correlation between cognitive level and the risk of injury was common in different gender, grade and caregivers educational level groups ( OR =0.87- 0.93 , all P <0.05). Sensitivity analyses yielded results consistent with the primary analysis. Conclusions The incidence of injuries among primary and secondary school students in Dongguan City is relatively high. Injury prevention knowledge level is negatively associated with injury occurrence risk, exhibiting an approximately linear dose response relationship.