Objective To improve the static and dynamic matching precision of OpenSim biomechanical models, and further improve the reliability for calculation of kinematic and kinetic parameters. Methods Model nonlinear scale based on kinematic experimental data was implemented by position adjustment, interpolation function calculation. For the verification of this method, two open dataset (GC3 and GC5) were used to build the nonlinear scaled models and calculate the limb lengths and joint reaction forces. The results were compared with those calculated by anatomical landmark scale (ALS) and linear scale method. Results The maximum discrepancies between limb length of nonlinear scaled model and actual model were 14.74 mm, which were in the range (4.0±13.8) mm reported by other literature. Marker errors of scale and inverse kinematic calculation could fulfill the requirement of OpenSim. As for calculated joint reaction forces, the root mean square errors (RMSEs) (GC3: 0.40 BW, GC5: 0.34 BW, BW was abbreviation of body weight) were smaller than those of anatomical landmark scale (ALS) (GC3: 0.64 BW) and OpenSim linear scale method (GC5: 0.40 BW). Besides, the results of Monte Carlo analysis indicated that, with the variation of initial positions of model markers, the range of joint reaction forces errors was smaller and limb lengths fluctuated within 5%. Conclusions The nonlinear scale method in this study is effective, and it can improve the efficiency of kinematic and kinetic modeling process and raise the precision of simulation results under current verification condition.