A stomatal conductance-photosynthesis-energy balance model developed by Leuning is used to estimate CO₂ and H₂O fluxes of temperate mixed forests. The canopy is divided into multiple layers, using both Gaussian integration and average- layer method. The fluxes are calculated layer by layer and then accumulated for the whole canopy. The model calculates CO₂ and H₂O exchange between atmosphere and canopy based on the input of meteorological variables. We compare the modeled fluxes with those measured by eddy covariance system, and analyze the influence of different layer-methods on the modeled fluxes. The energy balance ratios (EBR) of CBS, GDK and MMF sites are 0.76, 0.66 and 1.07, respectively, which is reasonable according to the international reported range (0.34-1.2). The results of the Gaussian integration method agree well with the eddy covariance fluxes, showing the diurnal CO₂ fluxes as bell shaped. The CO₂ flux (Fc) is negative and stable during the night, which suggests that forest ecosystems release CO₂ to the atmosphere. After sunrise, Fc increases gradually to positive values, peaking at noon, and decreases in the afternoon to the negative night values. The diurnal cycle of H₂O fluxes (LE) is similar to that of Fc, and the modeled LE agree well with the measured LE. The estimated fluxes are influenced by the different layer-methods. We compared CO₂ and H₂O fluxes from dividing the canopy into 2, 5, 10 and 20 layers averagely with the fluxes from Gaussian integration method. Larger numbers of layers produce lower LE and higher Fc. The fluxes from different layer-methods are sensitive to the vertical profiles of absorbed radiation, temperature, humidity and wind speed. Furthermore, the leaf photosynthesis responds nonlinearly to the radiation.