Informal settlements host around a quarter of the global population according to UN-Habitat. They exist in urban contexts all over the world, in various forms and typologies, dimensions, locations and by a range of names (squatter settlements, favelas, poblaciones, shacks, barrios bajos, bidonvilles, slums). While urban informality is more present in cities of the global south, housing informality and substandard living conditions can also be found in developed countries. These areas have common characteristics, including the deprived of access to safe water, acceptable sanitation, lack of health security and durable housing; in addition to being areas that are overcrowded and lack land tenure security. Such settlements are usually located in suburban areas, isolated from the core urban activities. The mapping of the urban form in such cases is a challenging task mainly due to their complexity, and their diverse and irregular morphology. Earth Observation plays a significant role in the mapping and monitoring of the extend, structure and expansion of such areas. Despite the increasing availability of data of very high resolution, standard methodological approaches usually fail offer high quality baseline data that can be used in urban surface and climate models, due to the aforementioned complexity (density of temporal buildings, mixing of materials used in the settlements, low height constructions). Here we present the first attempt to delineate the urban form of the slum of Mukuru in Nairobi, Kenya using Unoccupied Aerial System (UAS) data. Information from the slum, such as number of buildings and heights, density of structures, vegetation cover and height of high vegetation, digital surface model (DSM) and digital terrain model (DTM) are to our knowledge unavailable and consist of the main objectives of our approach. The above mentioned are usually the minimum spatial input requirements used in neighborhood-scale urban climate models such as the Surface Urban Energy and Water Balance Scheme (SUEWS). Data collection was performed in February 2021 covering an area of 4km2 using the Wingtra WingtraOne VTOL, a UAS system equipped with a fixed-lens 42 MP full-frame camera (Sony RX1R II) and accuracy of less than 2 cm using PPK. The images have been processed with the Wingtra application for the PPK corrections. The analysis of the imagery was run in Agisoft Metashape for the creation of the basic products: a) orthoimagery and b) DSM. The orthoimagery has been further analysed to derive a detailed five-classes (paved, buildings, high – low vegetation, base soil and water) landcover (LC) of Mukuru using a Random Forests classification algorithm developed using EnMAP toolbox in QGIS. The DSM product has in turn been exploited to derive a bare surface model (digital terrain model – DTM) following an approach based on a filtering method using moving window algorithm. DTM is the major input to create the normalized DSM (nDSM) as an intermediate step, in order to derive the heights of buildings and other objects (i.e., vegetation). The LC achieved an overall accuracy of 91.5%, with a class-wise accuracy of 1) Buildings at 90.16%, 2) Low and High Vegetation at 89.8%, 3) Bare Soil at 85% and 4) Water at 100%. In absence of GCPs from Mukuru slum, no validation was possible in the DSM and DTM products; GCP data collection was planned to run in summer 2021 but due to COVID19 situation and other safety reasons, up to now such data are not yet available. However, since the initial data has been corrected using PPK, we do not expect large errors in the elevation values of the landscape. Further analysis in the products of the building and vegetation heights shows over/underestimations of heights in areas with abraded changes in slopes such as the riverbanks. This is due to the methodology of the data collection process with the UAS, where, while the overlap was adequate in general, the use of a 3D grid for data collection would support the avoidance of errors in slopy areas. This study is the first one for the slum of Mukuru aiming at extracting the urban form and support the local microclimate modelling of the area using Urban Canopy Models (UCM) such as SUEWS.