The identification of settlement areas and their associated structures on early Holocene archaeological sites has long been an important topic in the study of the last groups of hunter-gatherers. In western France, shell middens constitute a very particular type of sites, which bear witness to very different domestic activities and are often interpreted as habitation sites. However, the residential structures at these sites are rare and never well known. The surroundings of these shell layers had never been excavated with modern protocols in Brittany until the field campaigns at the Beg-er-Vil site in Quiberon. The Quiberon Peninsula is located on the Atlantic coast, west of the Gulf of Morbihan, and extends 11 kilometres to the south. It lies in front of several islands: Téviec, Houat, Hoedic and Belle-Île-en-Mer. The Mesolithic site is located on the Beg-er-Vil promontory, which closes off a south-facing bay to the east, where the harbour of Port-Maria is located. It lies at the bottom of a small bay on the western flank of this rocky headland, which is only 5 metres above sea level. The Mesolithic settlement can be recognized on the beach promenade by a distinct layer of black earth laden with seashells, bone remains, ethics, and charcoal. It is located above a fossilised beach from the Eemian, about 3 metres above the highest sea level, and is covered by aeolian sands. The site was discovered in 1970 by G. Bernier and excavated over an area of 20 m² by O. Kayser between 1985 and 1988. Given the site's great scientific potential for understanding prehistoric societies on the French Atlantic coast, and due to the fact that it is directly threatened by storms and marine erosion, a new excavation program was launched between 2012 and 2018, managed by G. Marchand and C. Dupont. The calibration of the twenty selected dates (on short-lived samples) with Oxcal V. 4.3 suggests an occupation spanning between 6400 and 5927 cal. BC. The central sandy area of the site revealed a considerable number of structures in the form of hearths, pits, and vertical alignments of stones, interpreted as dwellings (or post-holes?). Due to the unique conservation of these type of archaeological discoveries in the region, the detailed study of these structures and the central area of the site is a crucial research topic for the study of the last hunter-gatherers’ populations in the region. The objectives of this paper are to present the archaeological data resulting from the excavation work in the sandy area of the Beg-er-Vil site, the detailed technological data supporting an interpretation of the original morphology of these structures, and the analysis of the micromorphological studies hinting at the formation of living floors. The results of the study show how the postholes, their organization, and their construction techniques, represent the last inorganic remains of a horizontal architecture, drawing an oval shape. These postholes are identified by empty spaces defined by a series of stones, mainly granite or quartz pebbles, placed on a vertical plan. In many cases, these were previously used as stone hearths. In the middle of this space lies a significant hearth structure, fitted with large granite stones, which according to C14 data was built at the beginning of the occupation phase. The series of vertical stones share some specific technological attributes, such as orientation, depth of implantation in the soil, and shape, that are significant to partially reconstruct this architecture even if many other elements are not (yet) known. With a width of 3.48 metres (east-west) and a length of 3.69 metres (north-south), this subcircular room covers an area of 10.08 m² in total. The 23 stone settings interpreted as stake wedges indicate that this structure was erected vertically by a scaffolding, probably made of plant material and consisting of parts with a diameter of 4 to 8 cm. To the north and south of the circular space, the absence of surviving stake wedges suggests at least three possible interpretations: that they were altered/destroyed by taphonomic processes (unlikely given the overall preservation of the site); that the poles of the armature at this level were set directly into the sediment, without stone wedges; or that they are in fact openings in the architecture. Given the conditions under which the site was buried, it is likely that these interruptions are genuine openings in the architecture. The geoarchaeological investigation of three soil-sedimentary sequences located within, near, and far from the building shows the spatial and temporal disparity of the archaeological stratigraphy. Micromorphological analysis confirms the presence of occupied Mesolithic soils associated with the formation of the shell layer. Several types of activity were detected in these soils. There are discharge-type inputs associated either with combustion activities (heated stone) or with the construction of structures (stake wedges, slab hearths) requiring burial of the underlying layer (rolled gravel beds). Relatively intense trampling can be observed in the upper part of the soil. Micro-remains are rare and consist mainly of micro-charcoals and food remains. Like the soils outside the structure, the soils inside do not appear to have undergone any special preparation; their formation seems to be exclusively due to natural processes. There are also no signs of matting, neither with plant nor with animal material. The soils show variations in wetting and drying, as do the soils outside the structures. The micromorphological investigation also provides clues for the temporal classification of this colonization. In the three sequences analysed, there was a decrease in the anthropogenic processes involved in the dynamics of the formation of the soil-sediment sequence, which could be explained by a period in which the area was not occupied by Mesolithic human groups.The function and uses of this structure cannot yet be deduced with absolute certainty. It might have been used for typical domestic activities on a daily basis, or for other types of activities related to food processing or crafting activities. The formation of anthropised soils suggests several occupation episodes related to the use of the structure. This study improves our understanding of the characterization of the archaeological remains of human groups on the coast, and of their ways of living.