The purposes of this study were to investigate the influence of changes in the subjective effort at the high intensity range on the ball speed in the pitching and to make clear its repeatability. Twenty male university baseball pitchers took part in this study as the subjects. Each subject randomly performed three sets of pitching at five different subjective efforts of 90, 92.5, 95, 97.5, and 100％ maximum. The ball speed of each trial was measured with the speed radar gun. After one week, the retest was carried out in the same procedure. The intraday (inter-set) and the interday intraclass correlation coefficients (ICCs) were calculated to examine the repeatability of the ball speed in each subjective effort. As the result, the ball speed tended to increase as the subjective effort was high. There were significant differences in the ball speed (all p<0.01) except the combinations of the subjective efforts next to each other. The relative ball speed that referred to the trial at the maximum effort as 100％ tended to be higher than each value of the subjective effort. This difference tended to decrease as the subjective effort was increased. There were significant differences among all combinations of the subjective efforts (all p<0.01). The intraday (inter-set) ICCs (0.90-0.96) and the interday ICCs (0.80-0.88) in each subjective effort were all significant (all p<0.01).