Evapotranspiration (ET) can be measured or estimated by using empirical models (FAO56 Penman-Monteith, Priestly-Taylor), direct measurement (eddy covariance, lysimeters), or indirectly (atmometer). The objective of this investigation was to measure and compare cumulative ET among eddy covariance (EC), lysimeters, atmometers, and empirical models [FAO56 Penman-Monteith (FAO56) and Priestly-Taylor (PT)]. The investigation was conducted at the Rocky Ford Turfgrass Research Center in Manhattan, KS. The study was initiated in July 2010. A weather station, three atmometers, and three microlysimeters were placed within the footprint of an eddy covariance system. The soil at the site was a Chase silt loam (fine, montmorillonitic, mesic, Aquic, Argiudolls). Microlysimeters were constructed from polyvinylchloride. The soil and vegetation in microlysimeters consisted of intact cores of tall fescue (Festuca arundinacea Schreb.) taken from the study site. In 2010, mean daily ET estimates were 5.51, 4.47, 3.68, 2.82, and 2.81 mm day^-1, for microlysimeters, PT, atmometers, FAO56, and EC, respectively. Therefore, ET from microlysimeters was nearly two times greater than estimates of ET from FAO56 and EC, 50% greater than from atmometers, and 23% greater than from PT, indicating significant differences in ET among techniques. These discrepancies could have important implications for ET studies in turfgrass. The fundamental mechanisms that caused the differences in ET among techniques are poorly understood and will require further investigation.