The purpose of this study was to clarify the factors that primarily affect the change-of-direction speed (CODS) in youth soccer players. Subjects were 70 youth soccer players. CODS were measured with a laser velocity doppler device. First, based on the 0m point, each subject was asked to start from the -10m position and change direction at a cutting line (5m from the 0m point) and returned to the 0m point as fast as possible. The section from the 0m to 4m points was defined as the approaching section, and the time within the section was defined as the approaching time (Tapp). The 1m section from the 4m point to the cutting line, where COD was made, was then defined as the cutting section, and the time within the section was defined as the cutting time (Tc). The section from the 4m point after the COD to the 0m point was defined as the accelerating section, and the time within the section was defined as the acceleration time (Tacc). Finally, the total of the times was defined as CODS. Subjects were classified into a fast group (FG) and a slow group (SG). CODS (p = 0.001) and Tc (p = 0.001) in FG was significantly faster than SG; however, no significant difference was observed in Tapp and Tacc. SG speed (p = 0.002) at the 4m point was significantly faster than FG (p = 0.015); however, FG deceleration and acceleration were larger than those of SG (p < 0.001). Meanwhile, peak points of deceleration (Pdec) and acceleration (Pacc) for SG were closer to the cutting line than those for FG (p ≦ 0.001). The results of this study suggest that Tc was affected by deceleration before COD. Therefore, it is suggested that soccer players are required to be good at deceleration as well as acceleration over short distances to improve their CODS.