Abstract: Working memory (WM) has been part of the research focus in interpreting.However, the complexity of both WM and interpreting makes their relationship not well explored by testing two comparable groups of college English majors at the beginning and end of an academic year.The present study tries to answer two questions:Does interpreting training improve WM abilities? Can WM abilities predict interpreting performance? With a close match in the pretest, the interpreting and control groups completed five WM tasks (verbal and non-verbal span tasks and an n-back task).The pre-test and post-test results revealed that:1) Compared with the control group, the interpreting group didn't show an advantage in WM; 2) For the interpreting group, English listening span and WM updating ability in the pre-test and in the post-test significantly predicted post-test interpreting performance.These findings help explain how WM functions in early-stage interpreting.