For students with diverse cultural or linguistic backgrounds who learn Japanese as a second language (i.e., JSL students), acquisition of Cognitive Academic Language Proficiency (CALP) and adaptation to classroom environment are pressing issues. These issues are also significant due to the increasing number of JSL students studying in Japan in recent years. To enable their study of music with verbal thought and verbal comprehension in music classes, it is necessary to establish an appropriate environment to facilitate their participation in class, not only through grasping the outline of a lesson in a context but also by understanding the verbal interaction between a teacher and students. This study aims to reveal the features of music teachers’ verbal interaction in music classes by focusing on the discrepancies between linguistic forms adopted them and their communicative intentions. Two lessons each at an elementary school and a junior high school in Japan were recorded; the collected data were analyzed by using the Classroom Interaction Analysis Sheet (CIAS) that encodes the verbal interaction made during classroom interaction. The study results suggested that the intended meanings of teachers’ verbal interaction varied, and that there were discrepancies between their linguistic forms and communicative intentions. The following patterns were observed: 1) teachers had no intention of asking questions despite using the interrogative form; 2) teachers used linguistic forms other than imperative sentences or directive statements although they intended to give orders or instructions. The findings imply that JSL students could not understand the intended meaning of teachers’ verbal interaction even if they had a foundational understanding of Japanese vocabulary and grammar. Thus, JSL students need support to deduce communicative intention in the classroom situation to enhance their learning.