predictive validity of the comprehensive basic science examination mean score for assessment of medical students' performance

background: medical education curriculum improvements can be a chieved by evaluating students' performance. medical students have to pass two undergraduate comprehensive examinations, basic science and preinternship, in iran. purpose: to measure validity of the students' mean score in comprehensive basic science exam (cbse) for predicting their performance in later curriculum phases. methods: this descriptive cross-sectional study was conducted on 95 (38 women and 55 men) guilan medical university students. their admission to the university was 81% by regional quota and 12% by shaheed and other organizations' share. they first enrolled in 1994 and were able to pass cbse at first try. data on gender, regional quota, and average grades of cbse, pc, and cpie were collected by a questionnaire. the calculations were done by spss package. results: the correlation coefficient between cbse and cpie mean scores (0.65) was higher than correlation coefficient between cbse and pc mean scores (0.49). the predictive validity of cbse average grade was significant for students' performance in cpie; however, the predictive validity of cbse mean scores for students' performance in pc was lower. conclusion: the students' mean score in cbse can be a good denominator for their further admission. we recommend further research to assess the predictive validity for each one of the basic courses.