Natural disasters, such as landslides, are a common problem in many places of the world, and studies about this topic have been increased since they lead to life and economic losses. Remote Sensing data can be used for mass movement studies, and for identification of landslides scars. So, this work identifies and characterizes landslides scars in Nova Friburgo – RJ, through digital processing satellite images of Landsat-5. To enhance the spectral and special quality, in order to facilitate the visual analysis and provide more information, the images were processed by removing clouds cover, selecting bands and obtaining color composition and performed a subtraction between bands. The Normalized Difference Vegetation Index (NDVI) was used to identify and mapping the landslides. A map of environmental fragility, based on Ross (1994) classification, was made to analyze the fragility in study area. Geological, geomorphologic and pedological data were researched to deepen the analysis. The results shows that the heavily rainfall in combination with multiple factors such as geology, geomorphology, soil saturation and high declivity, were the trigger to the 2011 landslide in the study area.