This study focused on the effect of land-use types and spatial scales on the evapotranspiration (ET) using remote sensing techniques. The processes included applying hybrid classification to generate a land-use map of northern Taiwan using Landsat-5 image in 1995; using DTM and the SEBAL model to calculate the parameters referred to ET and to compare the differences among different land-use types; and assessing the effects of 2 spatial scales (i.e., northern Taiwan and 7 watersheds within northern Taiwan) on ET using stepwise discriminated analysis. The result indicated that the area of northern Taiwan was classified into 7 land-use types (i.e., forest-land, building, farm-land, baring farm-land, water body, cloud, and shadow). The comparison of ET among different types showed that forest area had the highest value and then farm-land, water body, baring farm land, and building sequentially. As for the assessment of spatial scales on ET, the result pointed out that no matter which kind of spatial scales, the required parameters and the number of parameters for discriminating 5 land-use types (excluding cloud and shadow) were not the same, but NDVI and surface temperature were the most effective parameters. From above results, obviously ET varied with land-use types and spatial scales. Therefore, their effects on ET had to be considered in natural resource or water resource management.