The purpose of the present study was to analyse the results of the College Entrance Examination(CEE) and to examine the predictive validity of the CEE including the Scholastic Achievement Test for College Entrance(SATCE), the Achievement Scores in High Schools(ASHS), and the Essay Test for College Entrance(ETCE), which were done for selection of freshmen at Knagweon National University in 1986 and 1987. The subjects of the study were 2069 freshmen in 1986 and 2036 freshmen in 1987. The intercorrelation coefficients were computed to analyse the relationships among the three subtest scores including the SATCE, the ASHS, and the ETCE. Multiple regression analysis was done to examine the predictive validity of the CEE including the three subtests on the subjects' grade point averages of the college achievement for two years since the entrance of college as an external criterion. The major findings of the study are as follows : First, the interrelationships among the three subtest scores were very low : the correlation coefficients between the SATCE and the ASHS scores were -.04 and .06 ; between the SATCE and the ETCE scores were -.09 and -.06 ; between the ASHS and the ETCE scores were .02 and .06, respectively in 1986 and 1987. Therefore, the three subtests function independently as instruments for selection of college freshmen. Second, the multiple correlation coefficients of the three subtest scores with the subjects' grade point averages of the college achievement ranged from .322 to .484 in 1986 and from .312 to .410 in 1987 on the seven subcriteria of the subjects' grade point averages. They were significantly positive correlation coefficients at the level of .001. Therefore, the level of the predictive validity of the CEE was found to be a little below the medium. Third, among the three subtest scores, the ASHS made the greatest contribution to predict the college achievement scores for two years in both 1986 and 1987. It followed by the SATCE and the ETCE in that order. Fourth, there was a significant difference between the proportion of the variances of the college achievement scores which were accounted for by the ASHS scores, respectively in 1986 and 1987. It was found that the ASHS with the proprotion of 31.5% to the CEE in 1986 explained from 8.4% to 21.8% of variances of the college achievement scores and that the ASHS with the proportion of 40% to the CEE in 1987 explained from 6.6% to 11.3% of them on the seven subcriteria. Therefore, the ASHS with the proportion of 31.5% to the CEE in 1986 showed higher predictive efficiency than the ASHS with the proportion of 40% to the CEE in 1987. Finally, the CEE should be improved with its successive validation study. It is recommended that one should make good use of the ASHS as one of the major instruments for selection of the college freshmen.