The practical estimation of actual crop evapotranspiration (ELa) frequently involves calculating a reference crop evapotranspiration (ETref), and then applying suitable crop coefficients (K-c). Two standardized calculation procedures based on the Penman-Monteith equation are commonly used as ETref estimations (ETo for grass and ETr for alfalfa) across diverse environments. However, not much is known about the difference between ETo and ETr and their corresponding K-c values for maize in the arid environment of Northwest China with maize being the dominant crop. Estimates from the two procedures were comprehensively compared at both hourly and daily time steps for 2013 and 2014 growing seasons. The sum-of-hourly ETr and ETo (summed from hourly values over 24 h periods) values were averagely lower than their corresponding daily Er-r and ETo values (calculated from daily average meteorological data) by about 1% and 5% in 2013, and 2% and 7% in 2014, respectively. The ratios of ETr to ETo varied seasonally with a mean value slightly lower than the generalized value (1.35) for arid conditions, due primarily to the differences in responsiveness of both standardized PM equations to the trends and interations among input parameters. The K-c curves of maize based on grass-reference (K-co) and alfalfa-reference (K-cr) were experimentally derived using daily ETo and ETr combined with ETa measured by eddy covariance method. Results suggest the benefit and potentially improved accuracy for ETa estimation when applying alfalfa reference equation combined with the timestep-specific K-c curves under the arid conditions. These results provide an acceptable accuracy and precision for estimation of ETa of maize field use using the standardized Penman-Monteith equations, and offer the possibility of extrapolating new sets of K-co and K-cr to this region or other areas with similar climatic and crop management conditions. (c) 2017 Elsevier B.V. All rights reserved.