The arrival of Hurricane Mitch in Honduras in 1998 produced damages that were unprecedented in their wide-spread nature throughout the country. Estimations indicate that landslides damaged 70% of the road network in Honduras and nearly 1,000 fatalities. Abundant vegetation and deep roots are helpful to have stable soils and to limit the amount potential of a landslide. In this study we identify landslides after Hurricane Mitch and its vegetation recovery using landsat satellite images. The study area selected was in response to the priorization developed by local authorities. In El Cajon reservoir watershed 5 counties were selected and in Choluteca River watershed 14 counties. Multidate Landsat satellite images acquired previous (1997) and after (1999) the hurricane were pre-processed and used to derive Normalized Difference Vegetation Index (NDVI). The NDVI images were subtracted and a threshold was assigned to the resultant image to identify change areas. For the thresholding image generated, filtering techniques were applied in order to eliminate undesirable change pixels. When a proper threshold was selected, it was applied to the whole study area to identify landslides. To assess the accuracy of the landslide identification, ground truth was used. The vegetation recovery of these landslides identified was calculated for two periods, 1999-2003 and 2003-2010 in the two selected study areas obtaining a very high recovery for El Cajon reservoir watershed and low recovery for Choluteca river watershed area.