Earthquake on 26 December 2004 induces many small-scale landslides in southern of Thailand. In particular, people in many villages located within high hill terrains of Phang Nga province has suffered from this disaster. Therefore, the whole Phang Nga province located in Southern peninsular of Thailand was chosen as study area to obtain more accurate and reliable analytical results. In Phang Nga province, landslides have been recorded since 1987 by Nawavitphaisit (2005). Therefore not much attention has been paid in detail on these landslides. This study performed landslide susceptibility study using the landslide prediction model. Landslide susceptibility investigation was carried out in Phang Nga based on the work of Pantanahiran (1994). This technique relies on an inventory of past landslides. Remote sensing information i.e., aerial photographic images, satellite images through the digital enhancement technique was used for classification of landslide scars, deforested areas and bare land areas. The GIS applications were applied for evaluating landslide hazard area. In relation to causative factors, input parameters considered for calculation, adjusted aspect, slope, flow accumulation, flow direction, digital number of reflection (LANDSAT TM band 4), brightness and wetness. Geologic parameter has been added up in this landslide prediction model. With regard to rock types all over the country and from many trial and errors studying of past landslide in Thailand, we can categorize type of lithology into 4 classes of landslide probability unit value depending on degree of weathering susceptibility based on the work of Chotikasathien (2003).