Background: Chronic pain in children is a serious issue, therefore calling for effective prevention/intervention measures. This study aimed to evaluate the efficacy of an educational movie on pain knowledge in school children in general and on pain-related behaviours and pain intensity in those who are affected by chronic pain.Regarding those affected, the association between pain knowledge and intensity, as well as the potential mediating effect of pain-related behaviours, were investigated. Methods: Recruited from four schools, N = 381 students (51.7% female; M age = 11.4, SD = 0.95) participated, of which n = 108 reported chronic pain. Each school was randomly allocated to the intervention or control group (cluster-randomization). At two time points spaced 4-5 weeks apart, students provided information on their pain knowledge, pain-related behaviour (passive pain coping, pain-related disability, missed school days, medication use) and pain intensity. After the first assessment, students in the intervention group watched an educational movie. Multilevel linear models for all outcomes were calculated as well as a mediation analysis. Results: Pain knowledge increased significantly in the intervention group (β = 2.76 [95% CI 2.20, 3.31]). However, no significant time-by-group interactions were found for pain-related behaviour or pain intensity. The mediation model identified that the indirect effect of pain knowledge on pain intensity was mediated by pain-related behaviour (β = −0.18, p = 0.014 and β = 0.38, p < 0.001, respectively). Conclusions: Educational movies may be an effective tool for educating students about pain management. However, the knowledge gained may not be sufficient to improve pain behaviour and intensity overall. Significance: A 10-min educational movie on chronic pain management was tested in school children (N = 381). Following the intervention, knowledge of chronic pain concepts was statistically greater in the intervention group compared to the control group not watching the movie. Furthermore, a mediation model theoretically determined whether an association between pain knowledge and pain intensity exists, and whether this is explained by level of dysfunctional pain-related behaviorisms. Full mediation was confirmed.This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.