An investigation of evapotranspiration, vegetation quantity and composition, and depth to the water table below the land surface was made at three sites in two fallowed agricultural lots on the 15,800-hectare Tule Lake National Wildlife Refuge in northern California during the 2000 growing season. All three sites had been farmed during 1999, but were not irrigated since the 1999 growing season. Vegetation at the lot C1B and lot 6 stubble sites included weedy species and small grain plants. The lot 6 cover crop site supported a crop of cereal rye that had been planted during the previous winter. Percentage of coverage by live vegetation ranged from 0 to 43.2 percent at the lot C1B site, from approximately 0 to 63.2 percent at the lot 6 stubble site, and it was estimated to range from 0 to greater than 90 percent at the lot 6 cover crop site. Evapotranspiration was measured using the Bowen ratio energy balance technique and it was estimated using a model that was based on the Priestley-Taylor equation and a model that was based on reference evapotranspiration with grass as the reference crop. Total evapotranspiration during May to October varied little among the three evapotranspiration measurement sites, although the timing of evapotranspiration losses did vary among the sites. Total evapotranspiration from the lot C1B site was 426 millimeters, total evapotranspiration from the lot 6 stubble site was 444 millimeters, and total evapotranspiration from the lot 6 cover crop site was 435 millimeters. The months of May to July accounted for approximately 78 percent of the total evapotranspiration from the lot C1B site, approximately 63 percent of the evapotranspiration from the lot 6 stubble site, and approximately 86 percent of the total evapotranspiration from the lot 6 cover crop site. Estimated growing season precipitation accounted for 16 percent of the growing-season evapotranspiration at the lot C1B site and for 17 percent of the growing-season evapotranspiration at the lot 6 stubble and cover crop sites. The ratio of evapotranspiration rate to the reference evapotranspiration rate was strongly correlated with percentage of site coverage by vegetation at the lot C1B and lot 6 stubble sites (correlation coefficient = 0.95, sample size = 6), where percentage of site coverage was determined from quantitative vegetation surveys. It is concluded that evapotranspiration was mediated by the vegetation at all three sites, and that the differences in seasonal timing of evapotranspiration losses were caused by differences in timing of vegetation growth and development and senescence among the sites. Depth to the water table below the land surface at lot C1B ranged from 0.67 meters in early July to greater than 1.39 meters in late August. Depth to the water table at lot 6 ranged from 0.77 meter in late May to greater than 1.40 meters in late August.