This article evaluates the effectiveness of film analysis as a mediation strategy for teaching and learning Newton's laws in secondary education. An applied, quasi-experimental design was implemented in a nine-session teaching unit (blocks of ~80 minutes), structured around four criteria: careful observation of scenes, identification of physical elements, cinematic application (displacement, velocity, and acceleration), and linking to real-life situations. First-year high school students participated, comparing a group with film-based intervention to traditional teaching. The results show significant improvements in the post-test of the intervention group, with high effect sizes and substantial normalized gains. In addition, a marked reduction in misconceptions (e.g., force required to maintain MRU, confusion of action-reaction pairs) and greater transfer to unseen scenes were observed. The proposal showed high fidelity of implementation and increases in participation and engagement, especially among those with lower initial performance. Pedagogical implications, limitations (quasi-experimental, specific context, film selection), and recommendations for institutional scaling (curation of fragments, teacher training in film language and rubrics, Gantt planning) are discussed. The study supports the potential of cinema, combined with quantitative verification and critical discussion, to bridge the gap between theory and practice in dynamics