Student voice has become increasingly important in educational research at an international level. Research in Italy on school integration of students with disabilities has almost entirely left behind student voice. The very few researches based on student voice suggest that there is a mismatch between student and teacher voices when faced with the same situation. From a methodological outlook, the collection of data through students' samples implies some difficulties related to the fact that the study regards children and young students. The present issue becomes even more demanding when children with disabilities are taken into consideration. The aim of this research is to demonstrate the mismatch between the data produced by the teachers and the data produced by the students of the same class. These data are the basis for a study on the importance of student voice in research on school integration. The research was carried out as a multiple case study. Nineteen classes with at least one student with disabilities were involved. The teachers filled in a questionnaire with closed questions about school integration, later compared with the students' answers to the same questionnaire. A structured observation offered the opportunity to triangulate data thus adding new information on the reliability of the two points of view. The main results show that in all the classes there was at least one question with a significant difference between the students' and the teachers' answers; the triangulation of data through observation suggested that the students' perception was as close to observation as their teachers.