To solve a problem of a conventional listening skill development system that a learner is unable to objectively grasp improvement in his or her listening skill.

The listening skill is classified in steps according to the number of words. The number of words is increased stepwise according to (1) the skill to listening to one word, (2) the skill to listen to a sentence consisting of 2 words, and (3) the skill to listen to a sentence consisting of 3 words. This system enables the learner to objective grasp his or her listening skill, and always have a goal because of stepwise evaluation to continuously make efforts. Further, a learner's repeated voice is evaluated by speech recognition software, so the listening skill of the learner can accurately be evaluated, sound by sound.