This study investigates the effectiveness of applying task-based language teaching (TBLT) to medical interpreting course through a six-week quasi-experiment on 30 third-year clinical medicine majors. The classroom tasks used include interpreting doctor-patient conversations and teacher-student conversations in medical settings from English to Chinese and the other way around. All students participated in a recorded E-C interpreting test before and after the experiment respectively and the difference between the results of the two tests proves the effectiveness of TBLT in improving students’ medical interpreting capabilities. Some students received interviews about their opinion on TBLT and the result shows a majority of students hold a positive view on this pedagogy. The combined analysis of the tests and the interviews reveals the necessity to increase the input of relevant English terminology before starting each task and that students struggling with their other courses show little progress in the two interpreting tests. This study also offers suggestions on the course design of English for Medical Purposes.