Application of the traditional lithostratigraphic framework to subdivide theMiddle- and Upper-Quaternary locally-derived fine-grained deposits in theNetherlands is problematic. Deposits of many formations cannot bedistinguished from each other based on lithological characteristics andstratigraphie position alone. To overcome this problem, we present a new,well-defined lithostratigraphy for these deposits, based on detailedresearch in the central part of the Roer Valley Graben. This area containsan up to 35 m-thick sedimentary record of Middle- and Upper-Quaternary sand,loam and peat deposits. These have mainly been formed by aeolian andsmall-scale fluvial processes and have been preserved as a result oftectonic subsidence. The traditional lithostratigraphic subdivision of thesedeposits into three formations (Eindhoven Formation, Asten Formation andTwente Formation) was based on a combination of litho-, bio- andchronostratigraphic evidence and the presumed widespread presence of ahorizon of organic-rich interglacial sediments of Eemian age. To avoidintermingling of criteria regarding lithological characteristics, genesisand age, we now incorporate all fine-grained sediments into the new BoxtelFormation. The implications for the lithostratigraphic framework in otherparts of the country are explored and discussed. Eight lithostratigraphicmembers are introduced that describe the most characteristic parts of theformation. To fully illustrate the sedimentary sequence in the Roer ValleyGraben, two new members are defined here. The Best Member incorporatesalternating floodloam deposits and sandy aeolian deposits in the lower partof the Boxtel Formation. The Liempde Member includes reworked aeolian loessand sandy loess deposits (‘Brabant loam’) that occur in the upper part ofthe sedimentary sequence.

The original distribution pattern of Middle-Palaeolithic artefacts may be affected by tectonic movement, sedimentation and periglacial processes. This is e.g. the case in the coversand area of Limburg and Eastern Brabant (NL), where the occurrence of numerous finds in a SW-NE trending zone across the Roer Valley Graben is considered enigmatic. In order to elucidate the processes affecting the spatial distribution and the chance of recovery of such artefacts, we investigated a site in Nederweert. At this site, several Middle-Palaeolithic artefacts had been recovered earlier from unexpectedly shallow depths. A test pit profile and grain size analyses revealed that the shallow sediments at this site have been affected by intense, multi-phase cryoturbation, which has deformed the sand and loam layers and partially mixed them thoroughly. As a result, optically stimulated luminescence dating of these sediments yielded widely scattered single-aliquot equivalent dose distributions. Using a Finite Mixture Model (FMM), it was estimated that cryoturbation caused mixing of sediments deposited between 12 and 50 ka with sediment grains deposited between 60-150 ka. The latter material is probably the original context of the Middle-Paleolithic artefacts. Apparently, cryoturbation and potentially other periglacial processes have transported artefacts closer to the surface. Based on these results, we suggest that the occurrence of Middle-Palaeolithic artefacts is caused by (1) the tectonically-induced spatial distribution of layers of this age and (2) periglacial processes having caused migration of artefacts towards the surface. Although periglacial processes may facilitate finding Middle Palaeolithic artefacts, they may severely disturb the original context to such an extent that Middle Palaeolithic sites can no longer be identified. The results of this study form a basis for improving the Indicative Map of Archaeological Values that is used to predict the presence of archaeological sites. The insights gained are also relevant to other areas where Middle-Palaeolithic sites are affected by periglacial processes.