Observation of body movement of early childhood children listening to music has long been recognized as a key measure of development degree of music. It is integral part of standard music education, however, existing methodology depending on educators’ skill and experience present challenges for objectivity of evaluation. One potential solution is the integration of machine learning applied to feature value extracted from body movements. The author has applied full body motion capture and eye tracking and achieved fair accuracy of forecasting evaluation. In order to enhance machine learning accuracy, the author propose to introduce finger movement tracking to the platform. Finger movement incorporated through music generally accompanies linkage of hand and eye activity and children tend to coordinate full body movement with such hand and eye actions. In this study, the Meta gloves connected to the MVN as 3D motion capture system was used in order to analyze finger movements during musical expressions of hand-play songs performed by early childhood children. The author attempted to clarify the characteristics of finger movements through quantitative analysis of the third and second fingers, which were characteristic of the 19 types of finger movements acquired in musical expression. 3-year-old, 4-year-old, and 5-year-old children (n=86) at three facilities participated in the analysis of finger movements during musical expression in this study. A three-way analysis of variance was conducted on the moving distance, the moving average velocity, the moving average acceleration, and the moving smoothness of the third metacarpal bone and the adjacent second distal phalanx. As a result, it was found that for the third metacarpal bone, the moving distance and the moving average acceleration were larger for the melody (bright), and tended to be smaller for the melody (dark). On the other hand, it was found that the moving smoothness was larger for the melody (dark) than for the melody (bright). It was thought that the image in musical expression is formed based on the relationship that the third metacarpal supports the movement of the fingers and activates the movement of the second distal phalanx.As this study has shown, a more specific quantitative analysis of finger movements showing image of lyrics in hand-play song will contribute to the music education in early childhood because the extraction of feature quantities that contribute to the developmental process of musical expression in early childhood.