Hybrid implementation of pedagogical models (PMs) helps to overcome the limitations of a single pedagogical model (PM) when it comes to improving student learning outcomes in physical education (PE). Empirical research on hybridizations has grown substantially in recent years, so the purpose of this study was to conduct a systematic review on the effects and mechanisms of different hybridizations on students’ learning outcomes (i.e., motor, cognitive, affective, and social) in PE. Electronic databases, including ERIC, SCOPUS, EBSCO host, and Web of Science, were used to select intervention studies. After inclusion and exclusion criteria were applied, 17 high-quality studies, published in English peer-reviewed journals, were assessed. Results show that there were seven different hybrid models having impacts on students’ learning outcomes, which could be divided into four categories: (1) game performance and technical skills; (2) understanding of tactics and decision-making abilities; (3) motivation, autonomy, and confidence; (4) interpersonal skills, cooperative learning ability, and responsibility. Length of implementation and teachers’ familiarity were the main factors that limit the implementation on hybridizations. Future research should consider quasi-experiments with control groups of hybrids versus single models to figure out the advantages of the hybrid model over the single model; including more evidence from different schools, regions, and countries is necessary.