Abstract: Objective：To analyze and discuss the technical issues in serum creatinine assays and the role of reference measurement procedure for trueness verification. Methods：The trueness verification materials, quality control samples and the disturbed serum samples were measured by isotope dilution liquid chromatography tandem mass spectrometry (ID LC/MS/MS) method and 7 routine creatinine assays, including 4 enzyme methods and 3 Jaffe methods. The performance of each analytical system was assessed. Results：For the sample 201211 of trueness verification with low creatinine level, the relative biases were 13.2%, -5.2%, -4.3% and 10.2% for the 4 enzyme method respectively, and 5.0%, 28.0%, 29.5% for the 3 Jaffe methods respectively. For the sample 201212 of trueness verification with high creatinine level, the relative biases were 3.8%, 1.4%, -2% and 3.0% by 4 enzyme methods respectively, and 3.7%, 6.0% and 10.1% by 3 Jaffe methods are respectively. The average value of a patient serum with unspecific interference was 66.40 μmol/L in enzyme analysis systems and 102.29 μmol/L in Jaffe analysis systems. Conclusion：The specificity of Jaffe method is susceptible to be disturbed for the samples with low level of creatinine. The enzyme method may be affected by certain pharmaceuticals and reducing chemicals. The attention for the specificity and interference in creatinine measurement should be paid and the reference measurement procedure should be performed properly.