Abstract: Stomatal conductance is an important indicator of plant transpiration, and accurate quantification of stomatal conductance is significant for the study of the surface hydrological cycle. To improve the accuracy of stomatal conductance simulations, this paper uses two machine learning models, CatBoost（CAT） and Artificial Neural Network（ANN）, and a dataset consisting of Pinus edulis and Juniperus monosperma, to simulate stomatal conductance, and the simulation results are compared with the Ball-Berry model and Medlyn model. The input variables of net photosynthetic rate（An）, leaf surface carbon dioxide concentration（Cs）, relative humidity（RH）, saturated vapor pressure difference（VPD）, leaf temperature（TL）, and predawn leaf water potential（LWP） are designed as three modeling strategies in machine learning models. The input variables in strategy（1）are An, Cs, and RH; in Strategy（2）are An, Cs, and VPD; in Strategy（3）are An, Cs, RH, VPD, TL, and LWP. The results show that（1） Ball-Berry model and Medlyn model have similar simulation effects, with RMSE of 0.013 8and 0.013 9 mol/（m2·s）, respectively.（2） Compared with the Ball-Berry model and Medlyn model, the RMSE of the CAT model and ANN model under different input strategies decrease by 19.35%～45.65% and 26.90%～55.07%, respectively.（3） In the machine learning models, the simulation effect of strategy（3） is better than strategy（1） and（2）, and ANN is better than CAT. In strategy（3）, the RMSE of the ANN model is 36.70% and 38.54% higher than that of strategy（1） and（2）, respectively.（4） Stomatal conductance simulations for the entire dataset consisting of both plants under each model and strategy are consistent with the simulated patterns of stomatal conductance for Pinus edulis and Juniperus monosperma, respectively, with better results for Pinus edulis and then for Juniperus monosperma. These results indicate that the machine learning model（ANN model in particular） is more suitable for accurate simulation of the stomatal conductance of plants, and can provide a practical tool for estimating plant transpiration capacity and simulating agricultural hydrology.