DEM is needed for many branches of earth and environmental sciences such as control of erosion, flood planning, road construction, earth resource's recognition, and soil mapping. There are varieties of resources in generation of DEM, one of them is stereo satellite imagery. The objective of this research is the study of how to extract DEM from ASTER sensor stereo image. ASTER consists of three different subsystems: the Visible and Near-infrared (VNIR) and an additional backward telescope for stereo, the Shortwave Infrared (SWIR) and the Thermal Infrared (TIR). Each subsystem operates in a different spectral region, with its own telescopes, and is built by a different Japanese company. In addition, one more telescope is used to view backward in the near-infrared spectral band (band 3B) for stereoscopic capability. Images of Kashan area, which has been undulating topographic condition, were used. DEM extraction was based on the correlation algorithms between right image (3N) and left image (3B).The more correlation these tow images, the more precise DEM will be achieved. Finally, by correlation between tow images relative elevation attained. The results show that ASTER sensor, using along track imaging will attain the more precise DEM. The result accuracy of these images will be less than one pixel. This factor will improve the extraction of another production result by DEM.