This study aimed to was evaluating the performance of 30 methods to estimate reference evapotranspiration (ET0) to the city of Paranaíba, Brazil. The meteorological data was removed from National Institute of Meteorology, on the period of six year (March 2008 to February 2014). The method taken as standard was Penman-Monteith-FAO56 and the comparison of results was by the coefficients of determination (r²), coefficients “a” and “b” of the linear regressions, estimate of standard-error, Willmott’s index of agreement (d), Pearson correlation coefficient (r), and reliable coefficient (c). The better methods to ET0 estimate was: Penman-Original, Stephens-Stewart, Priestley-Taylor, Hicks-Hess, Turc, Liquid-Radiation, Thornthwaite-Modified, Temperature-Radiation, Penman-FAO24, Abtew and Camargo. The Camargo method should be preferred when only air temperatures data have. The methods Blaney-Criddle-FAO24 and Hamon should receive calibration for be utilized on the estimate of ET0 in Paranaíba city.