Objective To evaluate the effect of teaching evidence-based medicine (EBM) in postgraduates. Methods One hundred and thirteen postgraduates in the second year grade selecting an EBM course were included. The course lasted four weeks (twice a week, 3 hours each time and total 21 hours). The courses were delivered in combination with cases. The teaching effect was evaluated by requesting the students to present an evidence-based case report on the five steps of evidence-based practice. Each teacher assessed the reports independently. Results The mean score of 107 EBM case reports was 51.35±11.38. Distribution of the mean score was nearly normal. Score distribution of the items: the full mark rate was 97% for case depiction and 62% for evidence searching. Stratified analysis: ① Formulating answerable clinical questions: the full mark rate was 88% for clear statement of patient type, 89% for clear statement of intervention, 35% for control and 36% for outcome statement, respectively. ② Searching evidence: the full marks rate was 95% for selecting relevant database, 90% for searching term, 68% for searching strategy and 79% for searching result depiction, respectively. ③ Critical appraisal of evidence: the full mark rate of critical appraisal of the clinical importance and validity of evidence was very low (35% and 7%respectively). ④ Evidence applying: the full mark rate of intervention’s benefit and risk for individual patient and patient’s value was higher than that of applying evidence to patients and intervention feasibility (24% and 21%, 16% and 12%). ⑤ Evaluation: the full mark rate of clinical result evaluation was higher than that of performance of evidence-based practice (55% and 17%). Conclusions Teaching EBM for postgraduates was successful, we need to strengthen the practice of critical appraisal and applying evidence for patients in the future EBM course.