Abbott, W, Brickley, G, Smeeton, NJ, and Mills, S. Individualizing acceleration in English Premier League academy soccer players. J Strength Cond Res 32(12): 3512-3519, 2018-Global thresholds are typically used to band acceleration dependent on intensity. However, global thresholds do not account for variation in individual capacities, failing to quantify true intensity of acceleration. Previous research has investigated discrepancies in high-speed distance produced using global and individual speed thresholds, not yet investigated for acceleration. The current aim was to investigate discrepancies between global and individual thresholds when quantifying acceleration tasks. Acceleration data were recorded for 31 professional soccer players, using 10-Hz global positioning systems devices. Distances traveled performing low-, moderate-, and high-intensity acceleration were calculated for athletes using global and individual thresholds. Global acceleration thresholds for low-, moderate-, and high-intensity acceleration were classified as 1-2, 2-3, and >3 m·s, respectively, with individual thresholds classified as 25-50%, 50-75%, and >75% of maximum acceleration, respectively. Athletes were grouped low (LO), medium (ME), or high (HI) maximum accelerative capacity, determined using 3 maximal 40-m linear sprints. Two-way mixed-design analyses of variance were used to analyze differences in acceleration distances produced between analysis methods and athlete groups. No significant differences were identified between analysis methods for LO. For ME, no significant differences were demonstrated for low intensity. Moderate- and high-intensity acceleration distances were significantly higher for global compared with individual analysis method (p < 0.01). For HI, significantly higher acceleration distances were produced for all acceleration intensities using global thresholds (p < 0.01). Significant differences identified between analysis methods suggest practitioners must apply caution when using global thresholds. Global thresholds do not account for individual capacities and may provide an inaccurate representation of relative intensity of acceleration tasks.