As part of an intercomparison study on crop evapotranspiration (ETc), six methods for estimating ETc have been applied to maize field in themiddle Heihe River basin, China. The ETc was estimated by the soil water balance and Bowen ratio-energy balance methods while the Priestley–Taylor, Penman, Penman–Monteith and Hargreavesmethods were used for estimating the reference evapotranspiration (ET0). The results showed that the trend of ETc was very similar, while the differences were significant among the different methods. The variations of ETc were closely related to the LAI as well as to the meteorological features. The ETc for the Bowen ratio-energy balance, Penman, Penman–Monteith, soil water balance,Priestley–Taylor and Hargreaves methods totaled 777.75, 693.13, 618.34, 615.67, 560.31 and 552.07 mm, respectively, with the daily mean values for 5.26, 4.68, 4.18, 4.16, 3.79 and 3.73 mm day^-1.1. The Penman–Monteith method provided fairly good estimation of ETo as compared with the Priestley–Taylor, Penman, Hargreaves methods. By contrast with the Penman–Monteith method, the Bowen ratioenergy balance and Penman methods were 25.8% and 12.0% higher, while the Priestley–Taylor and Hargreaves methods were 9.4% and 10.7% lower, respectively. Therefore, the Hargreaves and Priestley–Taylor methods were the alternative ETc methods in arid regions of Northwest China.