Objective　To explore the influence of the selecting questions from national question database on the Pediatrics theoretical examination of clinical undergraduates under the mode of teaching-examination separation. Methods　to make an overall quantitative analysis of 304 test papers

selecting questions from national question database for the final pediatrics theoretical examination in the first term from 2017 to 2018

which was answered by two classes of clinical medicine students enrolled in 2014. Results　From the distribution of these two classes' scores

it can seen that the frequency distribution on both sides of the average is relatively symmetrical and all of them presented normal distribution

but they are all below average and good grades are fewer. They account for 5.29% and 9.15% of the total number respectively in each class. The number of students who failed of the two classes accounts for 20.53% and 16.99% respectively. The difficulty coefficients of the total volume are respectively 0.67 and 0.69

whose overall difficulty is moderate. Moreover

the test reliability coefficients of two classes are 0.764 0 and 0.768 9 respectively. The discrimination of the total volume of these two classes are all 0.22. There are some errors for the examination papers including in multiple choice options

repetitive questions

too many ques-tions in circulatory system and questions above student current levels. Conclusion　The result indicates that the testing questions show the difference of the students' ability and the reliability of the test through the overall quantitative analysis of the test. However

relatively low scores and a relatively large number of students failing the test indicate that students just stay in the old teaching and testing mode and they only master classroom knowledge. Therefore

they do not do self-learning well. Separation of teaching and testing could help to cultivate students' self-learning ability. At the same time

the quality of examination questions of the national question database and group questions should be constantly improved

which is the premise to ensure the objective reflection of students' learning ability.