BACKGROUND: This study aims to examine the relationship between functional movements and golf performance using the Golf Specific Functional Movement Screen (GSFMS). METHODS: This cross-sectional study included a total of 56 collegiate golfers (aged 20.89 ± 0.99 years, height of 174.55 ± 7.76 cm, and weight 68.48 ± 9.30 kg) who met the criteria, and were recruited from Hainan Normal University in June 2022. The participants' golf motor skills (1-yard putt, 10-yard putt, 25-yard chip, 130/100-yard set shot, driver, and 9-hole stroke play) were tested and the GSFMS (., pelvic tilt, pelvic rotation, and torso rotation) was used. RESULTS: There were significant weak or moderate correlations between the variables. Furthermore, a multiple linear regression analysis found that pelvic rotation and lower-body rotation abilities can significantly predict golf skill levels, which collectively explain 31.2% of the variance in golf skill levels among collegiate golfers (Adjusted R = 0.312, F = 2.663, < 0.05). Standardised β values indicate that pelvic rotation ( = 0.398) has a more substantial impact on golf skill levels than lower-body rotation ( = 0.315. CONCLUSIONS: This study found the weak to moderate correlations between the GSFMS and golf performance, and pelvic rotation and lower-body rotation abilities, thus predicting golf skills. Our findings provide novel insights into the relationship between functional abilities and comprehensive skill performance within the context of the Gray Cook's Movement Pyramid model, and provide theoretical support and practical reference for collegiate golf motor-skill learning and sports injury prevention.