The purpose of this study was to find a better evaluation method by comparison and analysis of the subdivision-score for preventive medicine with other subjects and total score. Among the 475 students who were second grade in 1983, 1984 and 1985, we analyzed the scores for all subjects at Yonsei University College of Medicine obtained by 443 students, with the exception of 32 students who had a temporary absence from school, failed or had been expelled. And we analyzed the score for preventive medicine of 162 students who were second grade in 1987 and 179 students who were second grade in 1988. Statistical analysis of the above was done using the correlation analysis, chi-square test and discriminating index. The results were as follows: 1. The correlations of the subdivision of preventive medicine in 1984, 1987 and 1988 were statistical significantly high(r=0.36-0.56). 2. The grades obtained for preventive medicine and for other subjects except pediatrics (clerkship) were not independent. 3. The discriminating indices that determined whether or not a pertinent evaluation was made were 0.42 in 1983, 0.52 in 1984 and 0.54 n 1985. These results were classified as excellent. These results suggest that the score of a subject is determined not by the characteristics of the subject, but by the amount of personal study.