The Tergra model simulates the daily course of water and energy flows through the soil-plant-atmosphere system and was intended for use with remotely sensed data. In its original form, the model is not well suited to estimating spatial patterns of latent heat flux (λE) in the Konza Prairie since the determination of canopy resistance requires knowledge of vegetation height, and the defined relationship between leaf water potential and rc is specific to C3 plants. The canopy resistance component of Tergra was replaced by a routine that includes the calculation of minimum canopy resistance (rcm) from the normalized difference vegetation index (NDVI) and stress adjustment factors for leaf water potential and vapor pressure deficit to determine actual canopy resistance (rc). The relationship between rc and leaf water potential is defined for both C3 and C4 plants, and total λE is obtained from the sum of the proportional contributions from these two vegetation classes. The modified Tergra model (Tergra-2) was tested using input and flux data collected at four First ISLSCP Field Experiment (FIFE) sites during three periods characterized by different soil moisture conditions. Tergra-2 was found to be a good simulator of λE and in most cases gave substantially better results than those obtained using the original model. The greatest inaccuracy using Tergra-2 occurred under extremely dry soil moisture conditions, whereas absolute errors for both models tended to increase around solar noon. Leaf water potential was the dominant stress factor affecting modeled rc. It was concluded that vapor pressure deficit and leaf water potential should not be regarded as completely independent factors affecting rc. A brief comparison of modeled and observed canopy temperatures is presented and discussed. ¿American Geophysical Union 1992