Provided is a coding training method using augmented reality as a method performed by a computing device including a camera, the coding training method including: a first step of obtaining coding training content; a second step of selecting one or more points included in the coding training content; a third step of obtaining an image photographed by the camera; a fourth step of recognizing the photographed image and matching one or more objects included in the photographed image with the one or more points; a fifth step of displaying an image and the matched one or more points; a sixth step of displaying a virtual robot on the photographed image; a seventh step of displaying one or more instruction blocks each including one or more instructions for controlling a movement of the virtual robot; an eighth step of arranging the one or more instruction blocks at the one or more points based on user input for the one or more instruction blocks displayed; a ninth step of controlling the virtual robot in order in which the one or more instruction blocks are arranged, moving the virtual robot along the one or more points, and performing instructions corresponding to the instruction blocks arranged at each point; and a tenth step of displaying the movement of the virtual robot according the control result.