Motor learning might reduce the likelihood of sports injuries by improving movement patterns. This review aimed to critically evaluate and summarize the effects of implicit (IL) and explicit (EL) motor learning on selected biomechanical variables while executing sport-specific tasks. PubMed, Embase, and Cinahl were searched according to PRISMA guidelines. The protocol was registered in PROSPERO (ID:​ CRD42023390982). Randomised controlled trials published before February 2024, with healthy participants (≥11 years) performing a sport-specific task were considered. Intervention evaluation, at least one kinematic or kinetic value as dependent variable, needed to be minimally one night post-practice. The changes in biomechanical outcomes were considered regarding the goal of the included study and their effect sizes were assessed. Methodological quality was based on the Revised Cochrane Risk-of-Bias tool. Of 18,639 records identified, 25 studies were included leading to 60 comparisons between IL and control (CTRL), EL and CTRL, or IL and EL. In total, the 1020 (520 male, 20.0 ± 3.7 years, 14–91 per study) participants playedvarious sports, from recreational to elite level. External and internal focus of attention, differential learning, analogy, non-linear and linear pedagogy, and observational learning were included.. An overall small positive effect size of IL compared to CTRL group changes was found (g = 0.45 [0.115 – 0.780], p = 0.01). Although premature, IL shows more potential than EL for improving biomechanics of sport-specific tasks. Therefore, practitioners may consider incorporating IL into injury prevention programs. Lastly, recommendations about strengthening research methodology of motor learning studies using biomechanical outcomes are provided.