Previous detailed studies of large rhyolite bodies propose that their elemental and isotopic characteristics were largely acquired in shallow crustal magma chambers. This model explains the common chemical and isotopic zonations of large volumes of rhyolites as well as the less common chemical and isotopic homogeneity of such bodies. We report an intermediate situation (the Estérel massif, southeast France) in which chemical variations contrast with Nd-isotope homogeneity. We thus infer that, in this case, large volumes of rhyolite resided for enough time in shallow magma chambers to develop chemical zonations through differentiation, but this process was not accompanied by crustal assimilation. The subordinate amount of mafic rocks cropping out in the Estérel probably evolved from basalt to trachyte through assimilation and fractional crystallization. The relatively radiogenic Nd-isotope signatures of the rhyolite compared with the Hercynian crust show that it cannot have been generated by partial melting of exposed basement rocks. Several geological similarities with large rhyolitic provinces could suggest that the rhyolite was purely mantle derived or, alternatively, generated by partial melting of an ad hoc crustal component. However, mineralogical, geochemical, and geodynamic connections between the Estérel rhyolite and the hypersolvus anorogenic granites of Corsica, as well as the extreme Nd-isotope homogeneity of the rhyolite, lead us to propose that the rhyolite was generated by mixing between mantle-derived magmas and a mafic lower crust. This scenario accounts for the relatively radiogenic Nd-isotope signatures of the rhyolite compared with the Hercynian crust. The good Nd-isotope homogeneity observed in the rhyolite implies that the mixing process, which occurred in the deep crust, was complete and provided a shallow magma chamber with isotopically and probably chemically homogeneous magmas.