Purpose: This study aimed to verify the difference between action anticipation and cortical neural activation for Kendo performance in the use of tools for professional kendo athlete and novice. Methods: Action anticipation tasks were performed on 15 kendo athletes and 15 novice group using temporal occlusion techniques. It is to accurately and quickly predict the final strike position by looking at the presented kendo strike videos (head, wrist and waist strike). In order to find out the differences in cortical neural activation according to action anticipation along with the task performance, motor-evoked potentials(MEPs) using TMS were measured. The dependent variables are predictive accuracy, response latency, and motor-evoked potentials, which are compared and analyzed between groups according to video length. Result: The analysis showed that the kendo athletes had an accurate and fast response latency compared to the novice. The comparison between groups according to video length was significantly accurate by the kendo athletes in sections 1 to 5, and was quickly performed by the kendo athletes in all video length conditions during the response latency time. In the comparison of the difference between the motor evoked potentials, kendo athletes were significantly higher in the 1st and 6th sections, and novice showed high cortical activation only in the 10th sections. Conclusion: As shown in the results, previous studies showed that kendo athletes’ excellent predictive abilities can also be seen in sports environments using tools, and behavioral predictive abilities are acquired through experience. The increased cortical activation of tool-use experts during behavior prediction demonstrated that mirror neurons do not simply follow motor matching between bodies, but also include movement of the tool, allowing mirror neural mechanisms to perform.