For long, the early Upper Palaeolithic has been at the background compared to the Magdalenian in the archaeological land-scape of Île-de-France and more generally of the Paris Basin, occupying a secondary place in academic research projects. This situation was however not due to a lack of discoveries insofar as since the origin of prehistoric science, almost in the middle of the 19th century, many amateurs and then rarer professionals have indeed brought to light and excavated some famous sites such as those of the Loing valley or the deposits of the Arcy-sur-Cure cavities. The problem came rather from the scientific orientation of most Palaeolithic archaeologists in the second half of the 20th century, responding to the siren calls of Magdalenian palethnology. Moreover, specialists of the Palaeolithic were perhaps not numerous enough to cover the 30,000 years of the Upper Palaeolithic. If A. Leroi-Gourhan and his team had not scientifically excavated some of the shelters of the Arcy-sur-Cure massif (Leroi-Gourhan et Leroi-Gourhan, 1964) and if B. Schnmider had not carried out some excavations on sites in the valley of Loing and its surroundings (La Pente-des-Brosses in Montigny-sur-Loing, Saint-Sulpice de Favières, Le deuxième Redan des Beauregards, Schmider and Roblin-Jouve, 2008), the records of pre-LGM archaeological deposits scientifically exploited in Île-de-France and more generally in the Paris Basin would be very thin. The few more recent preventive archaeology operations and rarer programmed excavations have not made it possible to restore for the moment the balance between Early Upper Palaeolithic and Magdalenian. This gap in the research effort is accompanied by a chronological gap: while certain periods like the Gravettian are well represented, others such as the Aurignacian suffer from a lack of well-preserved and therefore well-excavated deposits. Succeeding a collective research project on the Early Upper Palaeolithic in the south and centre of the Paris Basin which ended in 2005 (Bodu et al., 2013b), the excavations of the Ormesson site modestly contribute to show the interest of undertaking an intensive pro-gram of documentation, cartography and then excavations if we wish to fill in some of the gaps noted within this vast period. Thus, the open-air site of Les Bossats in Ormesson (Seine-et-Marne, France) is characterized by the presence in the same place of a succession of occupations ranging from the Middle Palaeolithic to the Iron Age but punctuated by various hiatuses. It unveils a long sedimentary sequence covering a very large part of the last Weichselian glaciation. A particular topography, an important sedimentation and strong but spatially limited erosive processes have allowed, in this sector of the Ormesson valley, the preservation of abundant evidence of the presence of human groups at least since the beginning of the Weichselian Pleniglacial. There is a great number of occupations for an open-air site, especially in the Paris Basin: in situ or reworked remains of Mousterian occupations older than 90-100 ka cal. b2k, a Mousterian level with discoid reduction covering an area of 800 m2 dated between 42 and 46 ka cal. b2k, a Châtelperronian complex spread over nearly 300 m2 and whose average dating is 42 ka cal. b2k, an occupation level not characterized in cultural terms and dated to around 32 ka cal. b2k (palaeo-fire?), a Gravettian locus dated 31 ka cal. b2k and a proper Solutrean camp bringing together 6 structures with very consistent dating around 23.5 ka cal. b2k. In this succession of chronocultures, a remar kable absence is that of the Aurignacian, even though there are evidence in the Paris Basin for decades. But at Ormesson, other absences must also be explained, as well as the presence of poorly represented Palaeolithic levels in the Paris Basin. In Ormesson, the taphonomic history of the dale and its slopes is essential for explaining presences and absences. The geoarchaeo-logical study of each occupation and the understanding of the morphosedimentary sequence, now based on a hundred ‘observation windows’, provide solid answers to the probable causes of certain chrono-cultural hiatuses and, conversely, to the existence of poorly documented Pleistocene archaeological assemblages for all northern France.