There is an increase of intensity and frequency of natural disasters around the world, especially in regards to mass movements (e.g., landslides), including in Brazil where several events occurred in recent years. Recent development of airborne remote data collection platforms and sensors such as unmanned aircrafts equipped with over-thecounter photographic sensors (drones) can, with a certain level of photogrammetric planning and data processing, be used to better understand these scenarios, with positive results in the identification, analysis and monitoring of areas susceptible to mass movement (e.g., landslide prone regions). The aim was (1) to produce an extensive and broad bibliographic review of related themes; (2) the usage of drones and ground control points (GCPs), by dGNSS technique, on a known geodynamic susceptible area (such as the Morro Doce slope, NW of São Paulo city); (3) a comparison by means of an appropriate algorithm of two registered 3D datasets (point clouds) generated from photogrammetric processing (SfM-MVS) in order to detect topographic change (e.g., a landslide). The first image set, acquired in 2017, contains 155 total images and 7 GCPs of <1 cm in XYZ precision. The second set, from 2019, comprises 484 images and 8 GCPs of <2cm in XY and <4cm in Z. In the face of different data acquisition arrangements (line of flight, height above ground, GCPs) and equipment, data processing and registration achieved the best possible quality by initial configuration analysis, camera model observations and automated cloud filtering and optimization (bundle block adjustment) via Python script. Then, the dense clouds, resulting from the MVS step, were filtered out of noise and above ground points and segmented to a common area (covering the instability features identified on the slope). Next, a multi-step dense cloud filtering approach took advantage of a specialized filtering algorithm (Cloth Simulation Filter), the calculated distances to a known reference surface/cloud (LiDAR-ALS data) and of a manual removal of points resulting on comparable point clouds free of objects (only ground points) with a registration error of 5cm measured on stable features of the slope (e.g., rocky outcrops). A 3D cloud to cloud comparison (M3C2 method) enabled the detection of changes which resulted in small significative changes for two bare soil landslide scars suggesting that these areas actually changed over the study time range, or that non ground points were still present on both clouds despite the careful filtering (since M3C2 considers cloud roughness). To take the photogrammetric data into account and improve the results, estimates of point precisions (M3C2-Precision Maps) were incorporated, resulting in no significant change detected and matching the local Civil Defense historic record for the time range. This result represents a more detailed historical change model when compared to the visual monitoring technique currently employed. Prevention is a major underpin in natural disaster management, thus the re...