Among the numerous Aurignacian open-air sites in southern France, “Régismont-le-Haut” (Poilhes, Hérault) is a rare example of a residential camp in which the explicit structure of a living space has been preserved. The camp covers a large and little disturbed area, at least within the two perpendicular paleochan- nels, which subdivide the site into two main areas.During the different excavation stages carried out at the site numerous hearth structures were discovered (n = 30), around which were distributed archaeological materials of varying densities, which indicated activity areas with a complementary function. Analysis of the nature and density of archaeological remains and also the proper characteristics of the so-called fireplaces and their association with discrete organised spaces (for some of which the existence of protective structures can be presumed) has identified two main loci, each located in a paleo- channel: the first contained several large polyvalent features and is interpreted as being a “domestic space”, while the second contained only specialised features (one of these most likely being a primary butchery area) and was interpreted as being a “working area”. All the archaeological materials were recovered within a single topographic surface. Moreover, connections between areas, established through the refitting of lithic pieces as well as the coherent nature and organisation of the different features within and between the two loci, provided strong support for a single occupation event, which was most likely a large seasonal residential campsite.Although the site can be clearly attributed to the Aurignacian sensu lato, its relative position within the Aurig- nacian chronological sequence is more difficult to establish. Given the absence of bone industry pieces (organic materials being poorly preserved at the site) and the fact that the personal ornaments recovered were exclusively made from shells (a ubiquitous feature within the Aurignacian), only the lithic industry can help us to place the site chronologically. Yet the nature of the lithic industry does not provide any clear clues allowing us to definitively place “Régismont-le-Haut” within the previously defined phases. The Late Aurignacian provides perhaps the best point of comparison, but it is difficult to ascertain this because an association with a particular Early Aurignacian facies, hardly known in the Mediterranean area, is also possible. Radiocarbon dates do not help to clarify the situation given the poor conservation of carbon in most of the samples treated. In addition, although a series of dates places the site between 31500 and 33500 cal BP, which corresponds to a Late Aurig- nacien stage, a recent set of dates suggest a time span between 35500 and 38000 cal BP. The gap between the two sets of dates is too large to be compatible with the geological data, making the hypothesis of several phases of occupation difficult to support. Moreover, the dated fireplaces belong to the same paleotopographic surface and both must have been rapidly covered by sediments so as to ensure their preservation; they could be sepa- rated by a few years, but not by hundreds or thousands of years. It is more likely that the gap between the two sets of dates is the result of taphonomic biases having influenced the carbon analysed. Regardless, the precise chronological placement of “Régismont-le-Haut” remains an unsolved problem.The chronological imprecision relating to the position of “Régismont-le-Haut” within the Aurignacian sequence does not, however, prevent us from exploring its importance with regard to Aurignacian human geography. Our first line of evidence to assist in this matter is raw-material provenances. The most informative material in this respect is lithic raw materials; while shell beads indicate that the Mediterranean coast was not unknown to the Aurignacians of “Régismont-le-Haut”. This area falls within the regional scope even if the coast was located farther away from the site at the time of occupation than is the case today. Although lithic raw materials were largely exploited on a regional scale, notably tertiary flints, extra-regional materials were exploited at distances of about 150 km as regards eastern sources, such as the Costières du Gard flints, and even farther away as regards the small assemblage of pieces stemming from the Northern Aquitaine basin (between 300 and 500 km when following the main valley corridors). How is it possible to explain the association of materials that were abandoned next to the same fireplace but stemmed from diametrically opposed regions? Several hypotheses can be proposed: - These objects were possibly obtained by exchange. This would mean that the group which stayed at “Régis- mont-le-Haut” and whose precise home territory remains unknown (groups from the Aquitanian region or from the Rhone valley meeting each other? Or groups in the Languedoc region exchanging with their neighbours?) used a toolkit that was, in part, obtained from other groups.- Or perhaps these objects result from direct acquisition, and if this is the case, reflect both very large scale group mobility and very long life cycles for the objects themselves, which accompanied individuals from the Northern Aquitania to the Rhone basin before coming to Languedoc.A third hypothesis can be advanced, which summarises the two previous ones: “Régismont-le-Haut” may have been a meeting point of several social groups, each having visited the Aquitanian basin and the Rhone basin respectively prior to their arrival at the site. In this case both the circulation and the exchange of objects, within and between different social units, would have taken place at the camp established at “Régismont-le-Haut”. However, the question remains: why in that place? The meeting of several social groups at “Régismont-le-Haut” could have been motivated by a successful encounter, by one group or another, during a hunting trip, providing one or several bison carcasses to treat and to consume; a hypothesis that is supported by the presence of bison remains at the site, most notably a nearly complete skull. The large quantity of meat and of potential raw materials (hide, bone) may have provided the impetus to establish a camp for a distinct period of time next to the kill site, and not far away from possibly rich hunting territories. This may support the idea of the presence of a rather large number of people, so the site could be interpreted as being a meeting place organised on the occasion of the capture of one or several animals, mobilising many members from a larger Aurignacian commu- nity from either side of the Atlantico-Mediterranean frontier while at the same time highlighting its permeability. Whether or not this scenario approximates reality, the data collected at “Régismont-le-Haut” are a remarkable illustration of the general impression provided by a majority of Aurignacian sites: vast territories in which emerged, through the fluidity of long-distance contacts, entrenched socially organised networks. In addition, it was possible to identify at “Régismont-le-Haut” the way in which the social organisation is mirrored by the spatial organisation of the camp occupation. “Régismont-le-Haut” appears to be a major open-air site for understanding the domestic organisation of an Aurignacian occupation, and moreover, based on the sheer size, it constitutes presently, and on the scale of the Early Upper Palaeolithic in general, at least in Western Europe, a remarkable example of a large residential camp site.