OBJECTIVE： To evaluate the uncertainty for the determination of pseudo-ginsenoside GQ （PGQ） concentration in human urine by LC-MS/MS. METHODS： The whole process of PGQ determination by LC-MS/MS in human urine was analyzed， and the uncertainty caused by repeatability， weighing， standard solution preparation， biological sample preparation， recovery， instrument precision and calibration curve fitting were evaluated. The combined uncertainty were calculated and expanded. RESULTS： The expanded uncertainty for low， medium， high （5.59， 48.07， 744.60 ng/ml） levels of PGQ in human urine were 0.58， 3.33， 50.62， respectively （P＝95％，k＝2）. CONCLUSIONS： The uncertainty of the PGQ determination in human urine by LC-MS/MS is mainly caused by recovery， repeatability of biological sample preparation and determination at low concentration； by sample preparation and recovery at medium and high concentration.