Automatic recognition of handwritten Chinese shorthand is a special issue of cha racter recognition subject. There is little literature on this subject. This pap er implements an on-line recognition system for HCSRS (handwritten Chinese shor thand recognition system) and provide experimental results for the recognition o f basic phonetic shorthands, shortforms, and linking shorthands. In this system, a structural recognition strategy based on the primitives of initial consonant symbol and rhyme symbol is applied. For efficient segmentation of primitives, a novel STSA (self-tuning segmentation algorithm) based on the segmentation-anal ysis-interaction configuration and corresponding knowledge sources is presented in this paper, which greatly improves the performance of the system. Theoretica l analysis and experimental results show that it is feasible to realize the simu ltaneous transformation from handwritten Chinese shorthands to written Chinese c haracters through computer.