Recently, have been evidenced in central/eastern French Pyrenees sub-marine polygenic breccias (Comus/Baixas Breccias), assigned to Upper Danian-Lower Selandian (P1c-P3) by means of planktonic foraminifera found either within their matrix, or within associated microrhythmic hemipelagites. These ante-Upper Eocene breccias, which are posterior to the HT-LP « Pyrenean » metamorphism (Mid.-Cretaceous in age and characterized by dipyre-bearing marbles and hornfelses) and to the Upper to Uppermost Cretaceous foldings, are only restricted to the Cretaceous orogenic axis of the range [Internal Metamorphic Zone (IMZ) and North-Pyrenean Zone (NPZ)]. They are dated in about 20 layers known from Mediterranean coast to Garonne valley. The breccias define in this part of Pyrenees a wide and long (more than 200 km) W-E trough (subdivided into several meridian palaeocanyons) inherited from former karstic topographies and separated by mountains with a steep topography, flanked to the South and the North of continental areas (covered by « Vitrollian » fluvio-lacustrine deposits). It was important to evidence if this marine breccia-filled « trough », Palaeocene in age, could extend westwards, West of Garonne, in Comminges/Barousse and Bigorre, where, laterally, the « Vitrollian » continental areas are replaced by outer-shelf marine sediments (clinoform carbonates), both covering the Sub-Pyrenean Zone (SPZ) and the High Primary Range (HPR) (Gavarnie-Mont-Perdu thrust sheet). In fact, the presence of those breccias has been already suggested (but without micropalaeontologic arguments) by Mattauer [in Choukroune, 1969 and 1976] in the Lourdes area (Bigorre). The topic of this paper is to characterize and to assign to the lower part of Palaeocene (63-59 Ma interval) several significant outcrops (St-Béat, Bramevaque/Troubat/Gembrié, Lortet, Medous/Bagnères-de-Bigorre and Lourdes/Pibeste) of these marine breccias (some of them previously used as black/yellow marbles called « Brèche romaine de St-Béat », « Portor des Pyrénées » or « Marbres de Medous ») recently identified from Garonne to Gave-de-Pau (fig. 1). Although quite poor in argillaceous hemipelagites, most of the breccias (which contain Mesozoic clasts) are now well dated by sections of « globigerinids » (= superfamily of Globerinacea) observed within their matrix. Other marine Palaeocene breccias also exist, more to the South (col de Gembre) along segments of the North-Pyrenean Fault, but they only rework Palaeozoic clasts. The « globigerinid » assemblage checked within all the Palaeocene breccias of Comminges/Bigorre includes, as more to the east, the following taxa: Globanomalina compressa, Gl. ehrenbergi, Gl. imitata, Parasubbotina varianta, P. variospira, Igorina pusilla, Morozovella angulata, M. praeangulata, Praemurica spiralis, Pr. inconstans and Woodringina hornestownensis. This assemblage is also laterally present within the marine carbonate sequences of the SPZ – HCR cover (« Lasseube Limestones » from the Nay/Pont Labau area, « Globigerinid-bearing Limestones » from the Gavarnie-Mont-Perdu thrust sheet), regions which are peripheric to the Pyrenean Lower/Mid. Cretaceous orogen (IMZ, NPZ) because exempt of major angular unconformity between Maastrichtian and Danian marine deposits (only a short gap of Lower/Lowermost Danian underlines the K/T boundary). On the contrary, the herein studied regions, belonging to this orogen, are characterized by a clear unconformity (both angular and cartographic) along a well-marked ravining surface inherited from erosional processes and karstification. The substratum of these breccias is strongly folded, cleaved and sometimes metamorphic and its younger formation seems to be Mid.– Cretaceous in age at least. Thus, it is very probable that the ante-Palaeocene unconformity seals compressional/transpressional structures (followed by emersions) assigned to the Uppermost Cretaceous phase (palinspastic transect, fig. 5). Danian/Selandian marine breccias and their already folded Mesozoic substratum are later tectonically reactived together by the « Pyrenean » compressions, Upper Eocene in age. If the elements of these breccias sometimes correspond to marbles induced by the Mid.-Cretaceous thermometamorphism (as around the famous « Etang de Lherz », more to the East, where lherzolites are also reworked in similar Danian/Selandian breccias), their matrix locally contain neogenic phyllites (never dipyre !) which could be related to a light (hydrothermal ?) post-breccia metamorphism. The clasts are generally angular, showing a very short transport from emerged steep topographies separating the different elementary canyons of the trough. The last problem is to determine the eventual westwards extension in the Bearn and Basque Pyrenees (fig. 6), particularly in the « Chaînons Béarnais » Zone which belonged to the North-Iberian palaeomargin (Iberian Plate) of the future range during Lower/Mid.-Cretaceous times. At this first level of micropalaeontologic investigations, it seems that several breccias (Lauriolle, Etchebar, Bosmendiette etc …), previously interpreted by several authors (synthesis in James and Canerot ) as Aptian and « diapiric » (collapse) breccias, should be assigned to marine Palaeocene deposits because containing (in their matrix and associated hemipelagites) Danian-Selandian planktonic foraminifera similar to the Comminges/Bigorre ones.