The purpose of this study was to clarify the step characteristics during the acceleration phase (0—30 m) of a 100-m sprint with reference to the step-type of sprinters. 59 male collegiate sprinters (season best time: 10.68±0.22 s) were recorded running in 100-m races using 10 high-speed video cameras (300 fps). The step variables, such as step-frequency (SF), step-length (SL), contact time, and flight time of each step in the acceleration phase (0—30 m) and maximum speed phase (30—60 m) were calculated. Cluster analysis was used to classify the subjects according to step-type as indicators for the ratio of SF and SL in the 30—60 m section. In addition, each of the step-type groups was divided into two sub-groups (good and poor sprinters) according to the mean speed in the 30—60 m section. The main results were as follows: (1) Sprinters were classifiable into 3 step-type groups; SL-type (n=22), SF-type (n=24), and Mid-type (n=13). (2) Among these groups, there were no differences in the 100-m race times and mean speeds during the 0—30, 30—60, and 60—100 m sections, although SL-type sprinters were taller and had a higher SLindex than SF-type sprinters. SF-type sprinters took a larger numbers of steps over the 100-m distance and showed a higher SFindex than SL-type and Mid-type sprinters. (3) In the 0—30 m and 30—60 m sections, the contact and flight times of SL-type sprinters were longer than those of SF-type sprinters. (4) In SL-type sprinters, the SF of good sprinters was higher than that of poor sprinters at the 5th step, and the SL of good sprinters was longer than that of poor sprinters from the 7th to 15th steps, and in the 30—60 m section. (5) SF-type sprinters showed no significant differences in the SL, but the SF of good sprinters was higher than that of poor sprinters from the 7th to 16th steps and in the 30—60 m section. These results indicate that there are differences in step characteristics during the acceleration phase according to step-type, and that the step characteristics during the acceleration phase may affect the acceleration ability needed to develop a maximum sprint speed. These findings could be useful for devising training methods for improvement of 100-m sprint performance according to step-type.