At the Niigata University School of Medicine, objective structured clinical examinations (OSCEs) were performed for the first time in 2001 for 92 fourth-year medical students. The average evaluation scores students received from instructors were summarized, and the differences between scores given by different instructors were examined. We found that practice methods for the medical interview and physical examination before OSCE and the question topics and evaluation methods of OSCE were appropriate, but scores on some items were extremely low. The standardization and objectivity of the evaluation were satisfactory, perhaps because one explanatory conference and two training conferences were held for instructors before OSCE. However, some questions tended to produce differences between instructors, as did some topics, especially in the medical interview. The scores with standardized patients and those by teachers were strongly correlated, but those with the former were lower than those by the latter.