Canopy conductance (g_c) is a key regulating factor of carbon, water and heat exchange between vegetation and atmosphere. Reliable and reasonable g_c estimation is of great significance for quantifying evapotranspiration (ET) mass and energy exchange at terrestrial surface. Based on the Jarvis model, a canopy conductance model of agroecosystem in an irrigated oasis, located in arid regions of Northwestern China, was formulated by using the time-piecewise functions of the response of leaf stomatal conductance (g_s) to environmental factors and Leaf Area Index (LAI). The developed gc model was tested with the calculated results derived from the inversion of the Penman-Monteith (PM) equation, in combination with observations of environmental variables and ET measured by the Eddy Covariance (EC) method, suggesting that the developed g_c model can provide reasonable prediction. In order to further assess the performance of the developed gc model, we consequently calculated ET under the conditions that LAI was larger than three, indicating that the estimation was in good agreement with the observations from EC method. It should be noted that the scaling leaf stomatal conductance to canopy conductance needs to take into account shelter factor (f_s), and the corresponding function relation with LAI is obtained by fitting. These results from our present study will provide a useful approach to quantifying the g_c of agroecosystems under the well-watered conditions in arid climatic areas, and then can improve the performance of ET estimation, which have important implications for well understanding the controlling mechanisms of plant on energy exchange and ET, and even for local water resources management.