Evapotranspiration is an important factor in estimating crops water use and then irigation schedule. Direct measurement of evapotranspiration is difficult since it is influenced by many factors. Estimation methods are developed for estimating evapotranspiration rate from meteorological data. One method which is recommended by FAO is Penman-Monteith Method (P-M). To evaluate whether this method could be accurately used in Lampung a comparison had been conducted with evaporation measurement on two climate stations in Lampung, Branti and Masgar with data set from 2006-2008. The result for Branti showed that observation data was lower than P-M for ET > 4 mm and higher for ET < 4; while for Masgar evaporation observation always higher than P-M. In general P-M was 1.09 times higher than observation in Branti and 0.89 lower in Masgar. Correlation coefficients between P-M and observation were low (r = 0.3 for Branti and r = 0.5 for Masgar). Two possible reasons for the disagrrement were first, there was an error in measuring water level on the evaporation pan, this showed by the fact that observed evaporation has low coefficient correlation with all meteorological data which have direct impact on evaporation (air temperature and humidity, wind speed and radiation); second, CROPWAT converted shunshine duration to be the radiation intensity with linear approach while field data showed that sunshine duration did not relate linearly with radiation intensity.