Objective It describes procedures for designing an analyzing a method comparison experiment for serum albumin(ALB) assay and determining the relative bias between two methods using split patient samples.Method According to the procedure described by the NCCL approved guideline: method comparison and bias estimation using patient samples; 40 patient samples were analyzed in 5 operating days, analyze each patient sample in duplicate using both BCG and BCP method. The duplicates were assessed for each method within the same run. The coefficient of correlation was calculated as well as the bia between the methods.Results In the patient′s serum sample, the bias between BCG and BCP methods were 1.8% at 50g/L or 5.5% at 40 g/L or 11.6% at 30 g/L for albumin. In the commercial quality control serum products, the bias between the two methods varied up to 36%.Conclusion In determination of the patient′s ALB, the bias between BCG and BCP goes up when the concentration of ALB is decreased. But the bias between the two methods varies up to 36% when the quality control serum is determined.