The purpose of the present study was to analyze effects of each factor of self-instructional training. Forty undergraduate students participated in the experiment. The task was mirror drawing in which subjects were required to trace figures quickly and accurately, referring to mirror images of pictures. Relative effects of factors of self-instructional training, i. e., external instruction, overt self-instruction, and covert self-instruction were examined. Measures taken were 1) speed and 2) accuracy in the task, 3) the standardized compound score which considered both speed and accuracy, and 4) the self-efficacy score. Results indicated that external instruction was significantly better than any other conditions in speed and the standardized compound score. Neither overt self-instruction nor covert self-instruction indicated any significant effects. The limited effects of self-instruction when solely used, and the effectiveness of external instruction to strengthen subjects' motivation were discussed.