A recently developed two-source energy balance model using directional radiometric surface temperature observations has been modified to use radiometric surface temperature observations at two substantially different view angles for computing surface fluxes without the need of an air temperature measurement. Near simultaneous dual-angle thermal infrared observations have been available form the along track scanning radiometer (ATSR) aborad the first European Remote Sensing Satellite (ERS-1) and therefore could be used operationally with this model. The required inputs to the new version of the model (2ANGLE_PT) are similar to the original model (1ANGLE_PT). They include directional radiometric surface temperature and its view angle, fractional vegetation cover or leaf area index, vegetation height and approximate leaf width, an estimate of the fraction of green or active vegetation, net radiation, and surface layer wind speed. Both versions of the model partition net radiation absorbed by the active vegetation into latent heat (LE) and sensible heat (H) flux according to the Priestley-Taylor (PT) approximation. The model- derived fluxes using multiple angle radiometric surface temperature observations collected from ground-based observations during the First ISLSCP Field Experiment (FIFE) conducted in the tall grass prairie in eastern Kansas during the summers of 1987 and 1989 are compared to measured surface fluxes. Differences between predicted and measured H using the 2ANGLE_PT model were significant while the original 1ANGLE_PT model yielded differences of 45-50 percent, on average. The 2ANGLE_PT model predictions of LE yielded differences of 25 percent with measured values while the 1ANGLE_PT model had differences of 15-20 percent on average. Using a simple technique to estimate daytime total LE from the 'instantaneous' values with the two versions of the model resulted in differences between 20 to 25 percent. Given the uncertainty that the surface flux measurements in FIFE were 20-40 percent, the performance of the models were marginal for H, but reasonable for LE. Since the 2ANGLE_PT model dies not require an air temperature observation, which cannot be measured for each satellite pixel and provided at the time of the satellite observation, this model has operational capabilities for computing regional LE with satellite observations having two substantially difference view angles, such as from the ATSR.