Landsat imagery can be used to establish spatial habitat suitability modeling for prediction habitat quality and evaluation potential habitat. However, the imagery has to limit information for specific habitat characteristics such as turtles' habitat. Chelodina mccordi is an endemic turtle from Roti Island but many previous studies have been done either biological or non-spatial studies. This study uses Landsat-8 OLI and TIRS as actual data and support with Landsat-5 TM as historical data, through a Geographic Information System (GIS). This study aims to create a spatial habitat suitability map model of C. mccordi and its mapping accuracy. The method that is used for modeling is overlaying indicative parameter maps use of logistic regression to presents the habitat suitability index (HSI). There are nine parameters used for modeling in this study, i.e. normalized difference water index (NDWI), land surface temperature (LST), slope, the topographic wetness index (TWI), distance from high canopy density, distance from settlement and agriculture, distance from freshwater, distance from the sea or salty, and distance from the street. We found three parameters that have strong contribution i.e. the NDWI, the LST, and the TWI. The result of spatial habitat suitability model of C. mccordi based on Landsat-8 OLI/TIRS and GIS analysis presents the value of mapping accuracy is about 75%.