Abstract: Objective　To provide the reference for traceability and comparability of calibration method of self-established detecting system through analyzing patient samples.   Methods　According to document EP9-A2 of Clinical and Laboratory Standards Institute (CLSI)，3 self-established detecting system FX，KH and FH (assessment methods) were calibrated by 4 different programs on a standard detecting system (reference method) composed of Hitachi 7180 biochemical analyzer and Roche reagents and calibrators (C.f.a.s). Based on 5 serum enzyme items including alanine aminotransferase(ALT)，aspartate aminotransferase(AST)，alkaline phosphatase(ALP)，gamma-glutamyltransferase(GGT) and lactic dehydrogenase(LDH) from 40 fresh samples，the differences between self-established detecting systems(Y) and the standard system(X) were analyzed，which included the correlation coefficient(r)，regression equation and the relative bias at medicine deciding level. Moreover，the coefficient of variation (CV) mean after calibration was compared with those of original detecting systems (Roche and Beckmann).   Results　The bias of 3 assessment methods had a obvious trend of changes. The results that calibrated by Japanese Certified Enzyme Reference Material(JCERM) and International Federation of Clinical Chemistry and Laboratory Medicine(IFCC) reference method were significantly better than those by calibrators from company and C.f.a.s. The CV mean of same enzyme item measured by JCERM and IFCC reference methods met or exceeded that of the original detecting system. Poor comparability of ALT，AST and ALP was shown by calibrators from company and C.f.a.s.  Conclusions　JCERM and IFCC reference methods are used to calibrate 5 enzyme items of the 3 self-established detecting systems with comparability and accuracy.

Key words: Enzyme, Comparative study, Assessment