In order for efficient post disaster management, it is important to develop a methodology to capture affected areas from remote sensing data. In this paper, the areas damaged by the 2006 Central Java earthquake are identified from the high-resolution satellite images. Based on the field survey for the spectral reflectance of surface materials, the image classification technique is applied to post-earthquake QuickBird images to detect the areas covered with bricks (damaged buildings) and the areas covered with roof tiles (undamaged buildings). The distribution of the building damage is evaluated by computing damage index from the classified images. The severely damaged areas are distributed not only in Bantul and Imogiri, located near the epicenter, but also in Gantiwarno, the southern part of Klaten located 20-30 km away from the epicenter. The distribution of the DI is compared with the damage statistics by the field survey. The result shows that the DI almost agrees with the severe damage ratio.