A leaf-scale combination model was modified to estimate canopy conductance (Gc) by introducing leaf area index (L). The modified model used only air temperature (T) and photosynthetically active radiation (Qp) as input variables, which can be measured directly. The Gc is the product of the potential canopy conductance (PSC), which is the L-normalized maximum Gc (GMax), and the relative degree of stomatal opening (RDO), which is the L multiplying by Gc normalized with GMax. The PSC is estimated from the maximum values of T, and the RDO is estimated from diurnal variation of Qp normalized by the maximum Qp. We applied the modified combination model to a winter wheat ecosystem in North China Plain. We also estimated canopy-scale latent heat flux (λEc) by coupling the combination model with the analogy of Ohm′s law in electrics. Simulated λEc for the winter wheat ecosystem was compared with that calculated from inverted Penman-Monteith equitation, and estimated λEc with observation by the eddy covariance system. The combination model yielded a quite well diurnal dynamics of Gc. Estimated λEc was in good agreement with that measured directly with the eddy covariance method. The slope and R2 of the line regression between the observed and simulated λEc are 0.705 and 0.79 (n = 1526), respectively.