The present study was to estimate pharmacokinetic parameters of metformin hydrochloride in  20 Chinese healthy volunteers with a limited sampling strategy (LSS), which will provide scientific data for  bioequivalence and clinical application.  A single dose of metformin was administrated to 20 healthy volunteers.  The concentration of metformin in whole blood was determined by validated high performance liquid chromatography (HPLC) method.  Multi-linear regression analysis was performed to establish a model to estimate AUC_{0−24 h} and C_max of metformin by LSS method.  The LSS models were validated by the Jackknife method.  The result indicated: the linearity relationship between AUC_{0−24 h} or C_max and single concentration point was poor.  Several models for metformin AUC_{0−24 h} or C_max estimation were better (r² > 0.9, P < 0.05).  Validation tests  indicated that most informative sampling points (C₂, C₆ for AUC_{0−24 h}, C_1.5, C₂ for C_max) provided accurate estimations of these parameters.  So, a multi-linear regression model for estimation pharmacokinetic parameters of metformin by using LSS method is feasible.