This study proposed a target prediction method based on the switch from ballistic eye movement to This study proposed a target prediction method based on the switch from ballistic eye movement to homing eye movement as a high speed and accuracy pointing method in eye-gaze input systems. The homing eye movement as a high speed and accuracy pointing method in eye-gaze input systems. The experiment was performed to verify the effectiveness of the prediction method. In this experiment, the experiment was performed to verify the effectiveness of the prediction method. In this experiment, the participants were required to carry out the pointing tasks. The experimental factors were the pointing participants were required to carry out the pointing tasks. The experimental factors were the pointing method (i.e., with or without prediction mode) , the target interval (i.e., 0.95 method (i.e., with or without prediction mode) , the target interval (i.e., 0.95 deg, 1.91 deg, 1.91 deg, or 2.86 deg, or 2.86 deg) , deg) , and the target size (i.e., 1.11 and the target size (i.e., 1.11 deg, 1.59 deg, 1.59 deg, or 2.07 deg, or 2.07 deg) . The dependent variables were the pointing time deg). The dependent variables were the pointing time and the percentage correct. As a result, the effectiveness of the prediction method was confirmed from and the percentage correct. As a result, the effectiveness of the prediction method was confirmed from the viewpoint of both pointing speed and accuracy. The prediction method could keep a short pointing the viewpoint of both pointing speed and accuracy. The prediction method could keep a short pointing time even if the target interval were small. Besides, more than 1.59 time even if the target interval were small. Besides, more than 1.59 deg target size is preferable to keep deg target size is preferable to keep a high pointing accuracy in the prediction method. a high pointing accuracy in the prediction method.