The report evaluates evapotranspiration at four grassland sites in eastern Washington: the Snively Basin and grass lysimeter sites on the Arid Lands Ecology Reserve, and the meadow and marsh sites on the Turnbull National Wildlife Refuge. A combination of the Bowen-ratio and Penman-Monteith methods were used to estimate evapotranspiration at the sites from May 30, 1990 to September 30, 1991. The Bowen-ratio method could be used to estimate latent-heat flux during only parts of the study period. Latent heat-flux values during these periods were used in the Penman-Monteith method to estimate canopy resistance. The daily average resistance values were used to recalculate latent-heat-flux with the Penman-Monteith method for all periods. Evapotranspiration estimates made with the two methods agreed within 3 percent over the period of study. However, for the grass lysimeter site, evapotranspiration estimates made with the Bowen-ratio method were only 41 percent of those made with weighing lysimeters. A water budget from August 20, 1990 to September 30, 1991 at the Snively Basin site estimated that 101 percent of the precipitation was returned to the atmosphere as evapotranspiration. Sixteen percent of the evapotranspiration occurred from October to February, while 76 percent occurred from March to July. April accounted for over 25 percent of the evapotranspiration for the water budget period.