For partitioning precipitation of tropical cyclone(TC) in China, an Objective Sy noptic Analysis Technique（OSAT）was used on six TCs affecting China in 2005. Th e results were compared with those results using subjective technique. It showed that some station observations caused by other weather systems or interaction b etween TC and other weather systems were included because the radius D1 of cir culation around maximum TC area was set too big. So the distinguish error is hig h, but missed unlikely. In order to reduce the distinguish error, a method of se tting D1 by realtime observations circulation radius and cloud and so on was proposed. Finallythe distinguish error was reduced, but missed a little more . The new result showed that rain station numbers partitioned by improved objectiv e technique could be less than that of subjective technique because the effect o f interaction between TC and other weather systems (cold front, west wind trough etc.) was not taken into account. Shortly, the accuracy of partition was improv ed after D1 was reset by integration of observation data and cloud image and s o on instead of intensity of TC. The precision of TC precipitation partitioned b y the objective method could be accepted in operation. It could be applied in op eration.