In a recent contribution to Basin Research, Gibson et al. (2007) advanced new thermochronological data from diierent massifs in the Pyrenees to argue for a 2 km of continuous post-orogenic exhumation since 30Ma. Gibson et al’s model (2007) is based on low-temperature thermochronometry from diierent sub-vertical pro¢les. This included apatite ¢ssion-track (AFT, most of them being previously published in Sinclair et al., 2005) and the ¢rst apatiteU-Th/He (AHe) data from the Pyrenees.Using Pecube software (Braun, 2003) they interpret their data as revealing a dramatic change in exhumation at about 30Ma from rapid ( 1.5mmyr ) to very low rates ( 0.03mmyr ) as earlier suggested by Fitzgerald et al. (1999) using the sole AFT thermochronometer from the same plutons. They considered such a decrease in exhumation rate to re£ect the transition into a post-orogenic state for the mountain belt. In addition, the presence of Oligocene to early Miocene AHe ages at the surface today is used by Gibson et al. (2007) to negate a possible reactivation of erosion during late Neogene as it has been inferred by geological and geomorphological data (Coney et al., 1996; Babault et al., 2005, 2006), and AFT data (Fitzgerald et al., 1999). Moreover Gibson et al. (2007) state that the preservation of peneplain surfaces at high altitude, as inferred by Babault etal. (2005), could not be possible. Our major concern with the study by Gibson et al. (2007) is that they implicitly consider that the rocks they sampled were exposed at the Earth’s surface only in the very recent past, that is at 0Ma. Gibson et al. (2007) recognize that their data do not allow determining the exhumation history since 30Ma. They cannot, therefore, rule out the possibility that the sampled rocks rose to the surface, or very near to the surface, in the late Miocene. Most of exhumation since 30^20Ma should be included in the process of peneplanation at high elevation that develops before the late Miocene as proposed by Babault etal. (2005). AFTandAHe data are in agreementwith such a scenario and cannot therefore be used as evidence to discard the well documented preservation of erosional surfaces in altitude in the Pyrenees. Finally, it is not surprising that Gibson et al. (2007) were not been able to determine a Plio-Quaternary increase of erosion, which corresponds mainly to a deep dissection of the former smoothed topography by £uvial network. Indeed, the amplitude of the £uvial incision, which never exceed 1000m, was not su⁄cient to alter the shape of the isotherms below the Pyrenean topography during the Plio-Pleistocene in a manner it imprints the AHe thermochronological data (see Braun, 2003). Firstly, contradictions in the estimates of rock exhumation byGibson etal. (2007) question the robustness of their conclusions.The authors state that (p. 331) ‘Since 29Ma, the amount of exhumation cannot have exceeded 2 km, at an average rate ofo0.03mmyr ’, which they followed by the opposite assertion: ‘The requirement of at least 2 km of rock exhumation . . . since 30-20Ma means . . . ’. Whatever the geothermal gradient they used to calculate the 2 km of crustal material removed since 30Ma, 0.03mmyr of exhumation during 30Ma correspond de¢nitely to 900m of erosion, not 2 km as stated! In essence, the thermochronological data documents the patterns of exhumation and rates up to a depth corresponding to a closure temperature, taken in Gibson etal. (2007) at 70 1C for AHe and near100 1C for AFTdata. Gibson et al. (2007) used geothermal gradients that vary from 20 to 35 1C km . Therefore these temperatures correspond to a range of crustal depths between 2 and 5 km.When the rocks are sampled at the Earth’s surface, as in the study by Gibson et al. (2007), there is a gap in the exhumation history of the rocks (between the closure depth and the surface) that does not imprint on AFT nor AHe data. Consequently, the details of the exhumation between the closure depth and the surface (de¢ned here as the post-orogenic exhumation) cannot be documented solely by thermochronological data. (1) Gibson et al. (2007) used Pecube software (Braun, 2003) to model the exhumation history of the Axial Zone (see their Fig. 4). There is an important assumption in their modelling of exhumation histories that is not clearly stated: the authors assumed that the ¢nal exhumation of Correspondence: J. Babault,Department ofGeology,Universitat AutoO noma de Barcelona, Bellaterra, Spain. E-mail: Julien. Babault@uab.es BasinResearch (2009) 21, 139–141, doi: 10.1111/j.1365-2117.2008.00377.x