The “Chemin de l’Évangile 3” (CE3) site near Gron is located on an alluvial terrace on the left bank of the Yonne river, in the Paris Basin. Excavations carried out between 2001 and 2006, made it possible to identify two archaeological levels, the earliest occupation (level I) taking place during the Solutrean and the latest one (level P) during the Last Glacial Maximum. A horse tooth found in the sands that underlie level I provided a 14C age of ± 24500 cal. BP. The palaeosoil, which contained level P, was dated by three samples to 18-19000 BP, i.e. between 24161 and 21750 cal. BP. Level P is the subject of this article. It was excavated over an area encompassing 217 m2 corresponding to the total area preserved at the site. The outstanding preservation of the remains, which were rapidly covered by river deposits, make it possible to carry out a complete analysis of the activities that took place at the site and to understand its functioning. The site was occupied during a rather short period of time in late winter or early spring. The remains of six animals (three horses, two reindeers and one bovine animal) were scattered across the site and in parts served as fuel to feed three fireplaces located in the northern part of the site. The vestigial remains were arranged around two fireplaces located 2 m away which concentrated most of the activities related to the processing of game as well as numerous activities related to flint working. These remains included waste stemming from the resharpening of burins, which were widely used in this area for the scraping and cutting of bone or wood. This sector appears to have played a central role in the organisation of the site. It also provided evidence of the preparation of flint projectiles and their possible replacement on a bone shaft. Ochre was brought to the site, then processed by crushing and grinding and was most likely used to produce an adhesive for micro-points. Large blanks were produced at peripheral locations at the site where they were used as bladelet cores or transformed into tools. In the southern part of the site mainly bone remains (horse and reindeer) were preserved and this area may have been a waste zone. All the fossil shells, over 60 remains in a poor state of preservation, were recovered in this area of the site. Given that the remains mirror the structuring of space, the question arises as to a possible organisation of this space around shelters. Bioturbation partially altered the distribution of the remains but did not totally disrupt their spatial organisation. The alignment of some lithic remains suggests wall effects. The duration of an occupation and the contemporaneity of the activities that took place are concepts which are particularly difficult to understand for the Palaeolithic periods. This period of time could correspond to the season of slaughter, which corresponds in Gron to the end of the winter and the early spring. The composition of the fauna does not show a discontinuity in time and the occupation could therefore have existed for a rather short time. The functional analysis reveals the complexity of the activities carried out : hunting, butchery, working of fresh and dry skins, exploitation of hard animal materials, and crushing/grinding of ochre. The use of bones as fuel makes it rather unlikely that distinct portions of carcasses were exported from the site. The processing of dry skins, if these are the skins of animals brought back to the site, takes some time and involves an occupation of several days. CE3 may be described as a "residential camp" in the sense of L. Binford (Binford, 1980). The organisation of space according to activity areas is known from Magdalenian sites such as Étiolles, Pincevent, and Verberie in the Paris basin. As is the case in CE3, domestic activities as well as distinct knapping activities were concentrated around hearths and links could be established with the peripheral areas. In contrast, CE3 encompassed a smaller area and may have been closer to what is called a "residential unit" that could be identified in the Upper Magdalenian sites. The lithic industry recovered at Gron is original compared to the known industries of the Last Glacial Maximum. From a technical point of view, distinct characteristic features of the Badegoulien in northern France are either lacking or rare in Gron, in particular the production of short flakes, used as blanks for so called small sidescapers (“raclettes”). The same is true for the transversal burins specific to the Badegoulien of the northern half of France, which are also rare in Gron. Equally, the production of micro-points at Gron has no counterparts so far and therefore is an utterly singular characteristics of this industry. Despite these peculiarities, Gron is not totally isolated among the material cultures of the Last Glacial Maximum in Western Europe. Many assemblages do not contain backed bladelets (Bodu et al., 2007) and comparative elements, which are often partial, can cause biases of interpretation. In the landscape of the beginning of the Last Maximum Glacial, it is possible that different human groups sharing the same technical system could be identified using peculiarities that were part of the history of each group. Each of these groups could then be identified by an identity signature, visible in hunting equipment or ornaments, and inserted in a larger cultural entity.