OBJECTIVE： To provide reference to promote the proper use of TCM decoction pieces. METHODS： According to Beijing TCM Decoction Piece Dispensing Regulation， stratified random sampling method was adopted to select a group of physicians in our hospital and conduct questionnaire survey about their knowledge of the prescription handling of 8 commonly used TCM decoction pieces including Scutellaria baicalensis， Angelica sinensis， Fritillary， Phellodendron chinense， Aconitum carmichaeli， Eucommia ulmoides， Ziziphus jujuba， Rhei Radix et Rhizoma. Then， the survey data was collected and analyzed. RESULTS： A total of 81 questionnaires were given out and all were returned with effective recovery rate of 100％. In terms of professional titles， associate chief physicians showed the lowest average error rates in prescription handling cognition （27.5％）， compared with physicians （35.8％） and attending doctors （37.0％）. They made most mistakes about Fritillary and E. ulmoides. To the contrary， both physicians and attending doctors made most mistakes about E. ulmoides. In the scope of working life， physicians from different groups showed similar average error rates in prescription handing recognition （31.3％-38.8％）. All of them made most mistakes about P. chinense and E. ulmoides. As for major， those who study western internal medicine shared larger average error rates （43.8％） than those who study internal medicine of TCM （36.8％）， traumatology and orthopedics of TCM （31.8％） and acupuncture （36.5％）. Each major showed the highest error rates in E. ulmoides. The order of 8 TCM decoction pieces from the most mistakes to the least mistakes made by physicians was E. ulmoides， P. chinense， Fritillary， Rhei Radix et Rhizoma， S. baicalensis， A. carmichaeli， Z. jujuba and A. sinensis. CONCLUSIONS： Physicians with different professional titles， working life and majors show deficiency in knowledge about the prescription handling of 8 commonly used TCM decoction pieces.  Generally， physicians show the worst understanding of E. ulmoides.