Evapotranspiration measurement is very important in research of energy and water flows in phytocenose, and for optimalized plant irrigation. Lysimeters are the most used deices, its principle is based upon change of weight of soil square and of vegetation during evaporation, or upon water balance. The size of lysimeter area has great effect on accuracy of measurement, and should be, according to WMO, at least 2 square meters (m2). For example, DWD lysimeter in Braunsweigh have 5 square meters (m2). CHMI measures maximum real evapotranspiration from standardized sward in Doksany. The lysimeter have dimensions of 2 m2 and resultant value of evapotranspiration is obtained from water balance equoation. At the lysimeter, amount both of naturally and manually added water, as well as amount of water passed through container is registered, also soil moisture is being measured. Taking into account difficulties with evapotranspiration measurement, indirect methods of evapotranspiration assessment are used more and more. These are so called virtual lysimeters, based upon mathematical models, which are using results of microclimate measurement in vegetation, or computing methods, that asre using standard meteorological data, or small water level evaporation measurements. CHMI is using Penman-Monteitha metod for estimation of evapotranspiration, while there appear to be perspective in using small water level measurement as improvement of this method. Based on experiences gained on Doksany observytory, EWM evaporimeter measurements (correlation coefficient r=0,892) are preferred to Class-A-Pan evaporimeter, at which there is necessary to make complex corrections for wind. For example, in vegetation season (April – September) of 2004, there was evaporation amount on EWM evaporimeter 620,8 mmm on Class-A-Pan 795,3 mm and 624,3 mm on lysimeter.