The major problem with high spatial resolution satellite images from Landsat is that imagery is not available very often (i.e. every 16 days or longer) and the coverage area is relatively small (swath width 185km), while images of lower spatial resolution from MODIS are available daily and one image covers a relatively large area (swath width 2,330km). This study considers the feasibility of applying various down-scaling methods to combine MODIS and Landsat imagery in order to obtain both high temporal and high spatial resolution. The Surface Energy Balance Algorithm for Land (SEBAL) was used to derive daily evapotranspiration (ET) distributions from Landsat and MODIS images. Two down-scaling procedures were evaluated: input down-scaling and output down-scaling. In each down-scaling scheme, disaggregated imagery was obtained by two different processes: subtraction and regression. The primary objective of this study was to investigate the effect of the different down-scaling schemes on the spatial distribution of SEBAL derived ET. We found that all of the four proposed down-scaling methodologies can generate reasonable spatial patterns of the disaggregated ET map. The results of this study show that output down-scaling with regression between images is the most preferred scheme and input down-scaling with subtraction is the least preferred scheme.