Although it is known that motor performance improves through motor skill learning, it is unclear whether the success rate representation affects the motor performance in association with the corticospinal excitability changes through the learning process. By using transcranial magnetic stimulation (TMS), we examined the changes in the corticospinal excitability of the engaged muscles (first dorsal interosseous, extensor and flexor carpi radialis, and anterior deltoid, FDI, ECR, FCR, AD muscles) before and after a short-term learning of a dynamic and coordinated upper limb movement. The TMS coil was placed on the motor hots pot of the FDI muscle. Nineteen healthy volunteers using their dominant hand were asked to hold a pen and dot (black ink) the center of an A4-size paper at which a small red ink dot was printed. The subjects performed the trials at 1 Hz, and were asked to put their hand higher than 100 mm apart from the paper surface trial-by-trial. All subjects were divided into two groups; one engaged motor learning without success rate (SR) representation (control group) and other with (SR group). With the control group, 50 trials per session, totally 20 sessions were performed in the motor learning sessions by using the same papers. With the SR group, on the other hand, the performance before learning (50 trials) was assessed before the experiment on a separate day, and a circle with a diameter of which the length was calculated by the success rate of 60 percents according to the performance before learning (50 points distance from the target), was printed around the target red dot in the motor learning sessions. The changes in the performance, electromyography (EMG), and motor evoked potentials (MEP) to TMS before and after motor learning were analyzed. The results showed that the error from the target significantly decreased through the learning in both groups, and the success rate was significantly smaller in the SR group compared with that in the control group. The EMG activity of FDI muscle was smaller, and that of AD muscle was larger in the SR group than those in the control group, while difference in the EMGs showed nether in the ECR nor FCR muscle. Through learning, interestingly, MEP in the FDI muscle was significantly larger in the SR group than that in the control group at resting state, but no difference in the MEP between the groups during voluntary movements (motor task). Also, the MEP/Background EMG ratio showed no difference between the groups. Our results suggest that, in the motor learning the performance can be further improved by representing the success rate of the motor task, and that the central motor command emphasizing on the proximal rather than distal muscle may contribute to the improved performance.