Abstract: Disposing the library functions call fast and efficiently is an important performance issue in the binary translation technology. One algorithm named JLSCL (jacket library and shortcut library) is presented to dispose the library functions classified, which is based on the dynamic binary translator integrated with static pre-translator. JLSCL can make use of the merit of dynamic translation and static translation, and make use of the convention of the function call on the target processor to reduce the redundant memory access. These make it more efficient that the original binary code runs on the target processor. The algorithm can flexibly switch between the library function and system call, and have applicability for the library function. The algorithm is verified in the binary translator system—digital bridge version 2. It can dispose the library function call successfully and efficiently, and its performance has been improved greatly.