In baseball batting, rotation around the long axis of the bat, know as " rolling ", has been observed. A batter who can attain a higher rolling speed before ball impact can achieve a higher rotation speed of the struck ball, which increases the ball’s flight distance. It has been suggested that batters who swing the bat with high nutation can attain a high rolling speed. The purpose of this study was to investigate the effect of instruction aimed at increasing the rolling speed in baseball batting. Ten batters performed tee-stand batting under 2 conditions: a usual swing (CON1), and a swing after being instructed to position the bat vertically, and then swing by rapidly lowering the bat head (CON2). The three-dimensional motion of the bat was measured using a small accelerometer and gyro sensor attached to the grip-end of the bat. This sensor was able to measure the swing speed, swing time, rolling speed, swing angle (angle between the bat head velocity vector and the horizontal plane) and vertical angle of the bat (angle between the long axis of the bat and the horizontal plane) before ball impact and the swing trajectory from the start of the swing until ball impact. The rolling speeds employed were 726°/s (CON1) and 854°/s (CON2). The rolling speed for CON2 was significantly higher than that for CON1 (p <0.05). On the other hand, there was no evident difference in swing speed between CON1 (30.1 m/s) and CON2 (30.2 m/s), nor were there any differences in other swing parameters before ball impact. Batters who swung the bat at a high nutation speed in response to instruction increased the rolling speed, but those who were unable to change the swing trajectory and nutation speed failed to change the rolling speed. These results indicate that batters increase the rolling speed without changing swing parameters such as swing speed, swing time and the vertical angle of the bat in response to the above instruction.