Abstract: OBJECTIVE To analyze the measurement uncertainty in determination of levodopa methyl ester hydrochloride in plasma by LC-MS/MS. METHODS Thermo BDS HYPERSIL C₁₈ column(4.6 mm×100 mm, 2.4 μm) was used, the mobile phase consisted of methanol and ammonium acetate (10 mmol·L^-1, pH=5), using?gradient?elution, and the flow rate was 0.4 mL·min^-1. Mass spectrometry was performed in multiple reaction monitoring (MRM) mode, with m/z 212.4→152.3 for levodopa methyl ester hydrochloride and m/z of 166.4→148.3 for internal standard. A mathematical model was established to analyze the source of uncertainty. The various factors influencing the uncertainty in the whole process of determination, including precision, weighing, solution preparation, plasma containing drug preparation, plasma extraction, the apparatus and calibration curves fitting, were all analyzed and estimated. The uncertainty and synthetic uncertainty of each variable were calculated. The expanded uncertainty was analyzed with all the components. RESULTS When the confidence probability P=95%, the expanded uncertainty for low, medium and high(0.4, 4, 32 ng·mL^-1) concentration of zidovudine were 0.10, 0.23, 1.62 ng·mL^-1，respectively. CONCLUSION This method is suitable for the uncertainty evaluation for the determination of levodopa methyl ester hydrochloride concentration in human plasma by LC-MS/MS, and provides reference for uncertainty evaluation of complex biological matrix analysis.