Purpose <p>Athletes’ safety and performance enhancement largely depend on the timely prevention of sports injuries. Artificial Neural Networks (ANN) have recently become valuable tools due to their efficiency and interpretability. Despite the promising potential of ANNs in preventing sports injuries, their validity and prevention power in a wide range of sports still need to be discovered. This study aimed to address this research gap and evaluate the effectiveness of ANNs in preventing sports injuries.</p> Methods <p>This research adhered to the rigorous guidelines outlined in the Preferred Reporting Items for Systematic Reviews (PRISMA) statement. To identify relevant studies, we conducted a comprehensive search across electronic databases, including PubMed, Scopus, Web of Science, and IEEE. Our review encompassed articles published from 2013 to 2023, ensuring a thorough and up-to-date analysis.</p> Results <p>This review included 28 studies that examined how well ANNs can predict the risk of sports injuries based on factors, such as performance, previous injuries, personal data, and biomechanics. In these studies, ANN types and different risk factors were used to predict the probability of injury in basketball (<i>n</i> = 9), soccer (<i>n</i> = 7), handball (<i>n</i> = 5), volleyball (<i>n</i> = 2), Australian football (<i>n</i> = 2), rugby (<i>n</i> = 1), baseball (<i>n</i> = 1), beach volleyball (<i>n</i> = 1), and running (<i>n</i> = 1) athletes. The accuracy of the multi-layer perception models ranged from a minimum of 51.21% to a maximum of 98%.</p> Conclusion <p>The findings of this study underscore the potential of ANNs as practical tools for sports injury prevention. The implications of this research suggest a paradigm shift toward more data-driven and personalized approaches. To further validate and refine these models, future research should focus on longitudinal studies, real-world implementations, and the inclusion of diverse athlete populations.</p>