Abstract: Objective　To compare methodologically and evaluate the bias of turbidimetric assay and particle-enhanced nephelometric assay for serum cystatin C(Cys C). Methods　According to the National Committee for Clinical Laboratory Standards (NCCLS) EP9-A2 document，40 serum samples were determined in duplicate measurement by Roche MODULAR P800 automatic biochemical analyzer and Siemens BN-Ⅱ specific protein analyzer， respectively. The data were used to check outliers，linear regression was analyzed，and the predictive bias and confidence interval（CI） were calculated. Results　The linear regression equation was Y=1.216X-0.139，r 2=0.978. When the Cys C concentrations were 0.53 mg/L and 0.95 mg/L，the acceptable biases were in the range of 95%CI of predictive bias. However， when the Cys C concentration  was 7.85 mg/L，the acceptable bias was＜the low limit of 95%CIof predictive bias. Conclusions　The results of the 2 detection systems have linear correlation. The predictive bias of measured results between the 2 detection systems is acceptable except the high concentration.

Key words: Cystatin C, Turbidimetric assay, Particle-enhanced nephelometric assay, Bias