Surface evapotranspiration is an important link for maintaining the water and heat balance of the earth. The SEBAL model is a remote sensing evapotranspiration estimation model used widely in the evapotranspiration research field. Surface albedo is not only a significant factor for the surface energy balance but is a vital input parameter in the SEBAL model. Different surface albedo computing methods will affect the results of the SEBAL model. In this study, we used two surface albedo computing methods, which were proposed by Smith Ronald and Liang Shunlin, respectively, to calculate the surface albedo for the Sangong River Basin in Xinjiang. Then, the calculated surface albedo was input into the SEBAL model to obtain a diary surface evapotranspiration. By analyzing two kinds of evapotranspiration and surface albedo results, we were able to draw several conclusions. The diary evapotranspiration results calculated from the two kinds of surface albedo methods and the SEBAL model fit well with the measured values. The smallest coefficient of determination was 0.75. The evapotranspiration results calculated using Smith's surface albedo showed a higher correlation with the measured diary evapotranspiration value than that of Liang's surface albedo. We compared the two diary evapotranspiration results using RMSE and other indicators, and we found that the evapotranspiration results from Smith's albedo method was more accurate than Liang's. The average diary evapotranspiration using Smith's surface albedo was higher than that of Liang's albedo; the difference was greater in the summer than in the other seasons tested (0.64 mm/d). The difference in the other seasons was smaller (0.2 mm/d). An additional comparison of the two kinds of average surface albedo revealed that the average surface albedo calculated by Smith's method was less than that of Liang's method for 6 periods.