Objective To provide data support to test question used in medical examination of basic medicine in order to meet the requirement of competency-oriented medical education assessment through comparing the item parameters between integrated items and non-integrated items in basic medicine. Methods Two item formats, integrated item and non-integrated item, on the same basic medicine knowledge point were used to in a large-scale clinical medicine examination. Fifteen knowledge points were chosen and the cognitive level, assessment point, difficulty and discrimination of different item formats were compared and analyzed, with significance test using pared-samples Wilcoxon rank sum test for difficulty and pared-samples T test for discrimination. Results The cognitive level of non-integrated items were memory and comprehensive, while the cognitive levels of integrated items were application of basic medicine knowledge points. The average assessment point number of integrated items increased 2.8 than that of non-integrated items. The difficulty of non-integrated items was 0.77(0.59,0.92), and that of integrated items was 0.75 (0.69,0.86). There was no significant difference between the two item formats (P=0.100). The average discrimination of non-integrated items was (0.26±0.11), and that of integrated items was (0.31±0.10). The discrimination of integrated items was significantly higher than that of non-integrated items (P=0.015). Conclusions The integrated item format meets the requirement of competency-oriented medical education and reform, complies with the tendency of medical examination assessment, and improves the quality of test questions of basic medicine.