The purpose of this study was to investigate cross-sectionally the development of running speed, step frequency, step length, running form, and maximal anaerobic power (MAnP) in high school boys. Subjects were 134 high school boys ranging from fifteen to seventeen years of age. They participated in power tests including the 50 m dash, maximal bicycle pedalling, vertical jump and long jump. They were videotaped during the 50 m dash to analyze their running form.Variables analyzed were: 1) 50 m running time, running speed, step frequency and step length, 2)step time, support time and non-support time, 3) path of the whole body center of gravity (CG),4)maximal anaerobic power, vertical jump, and long jump. The results were summarized as follows; 1. The running time for 50 m decreased from 7.44 s to 7;14 s with age. Running speed increased from 7.70 m/s to 8.00 m/s with age. Step frequency increased, while step length and step length/body height ratio showed no significant change. The increase in running speed resulted, therefore, from the increase in step frequency rather than the step length. 2. MAnP increased from 766.0 W to 823.6 W with age, but no noticeable change occurred in the MAnP/body weight. 3. At each age, significant correlations were found between step frequency, step length,and step length/body height except step length/body height of fifteen year old boys. Furthermore, running speed and variables for physical fitness such as MAnP showed significant correlations. Running speed and MAnP/body weight showed a significant correlation for each age group. 4. The relationship between running speed and the CG motion indicated that it was important to shorten support time, increase step frequency and reduce the bouncing motion of the CG in order to improve running speed. The relationship between running speed and the motion of the lower limbs suggested that motion of the toe relative to the hip during support phase, forward motion of the free leg and backward swing motion of the toe prior to foot contact at high speed help to improve running speed. 5. Based on the averages and the standard deviations of variables examined for each age group, running performance was classified into five categories. Averages of step frequency, step length, step length/body height, MAnP/body weight, vertical jump and long jump for the categories were listed in Table 6. This will be useful to evaluate the running performance of students and teach sprint running in physical education classes.