Preparation for the introduction of a long-term care insurance system is under way with various trial schemes being carried out. This paper determines the effectiveness of the interviews carried out in Ube City in the fiscal year 1998 as a part of the scheme of deciding the long-term care grade of a patient. Altogether we studied cases of 46 patients. One hundred and thirty seven health care professionals conducted the interview using the questionnaire decided upon in the fiscal year 1997. Each patient met 2-5 interviewers Eighty two interviewers knew the patients whom they interviewed. There was no illness on which the conclusions drawn by different interviewers from all 73 questions were matched, and only about 52% of the patients were given the same long-term care grade by different interviewers. There were two main causes for this outcome. First each interviewer interpreted questionnaires. Secondly, the interview was mainly done through listening to the patients rather than the patients themselves filling in the questionnaires. Therefore, in order to improve the standard of assessing the long-term care grade for each patient, it is recommended to put more emphasis on interviewer training and to give the interviewers more experience.