Paris Basin Tertiary formations contain large deposits of lacustrine limestones. The limestones have a low clastic content and many secondary dewatering and pedogenic-paludine fabrics indicating deposition in shallow environments. These lacustrine limestones commonly contain cherts that crosscut sedimentary structures. The silicified zones may be pervasive and retain the structure and dull aspect of the limestone or form irregularly shaped translucent nodules. Quartz is almost the only silica phase present in the cherts. Two main types of silicification occur together: (1) voids partly or entirely filled with quartz, and (2) limestone matrix that has been replaced by microcrystalline quartz with preservation of most of the primary limestone fabric. There is a systematic relationship between silicification and high-porosity zones. The replacement of the limestone matrix by quartz is directly connected to voids infilled with quartz. Because the limestones are pure, without clayey layers, the silica must have come from other formations (overlying sands and soils) and been introduced by groundwater flow. In view of the weak solubility of silica in surficial waters, substantial groundwater flow is needed to supply the silica precipitated from the solution. This explains the observed relationships between voids and silicification. A coupled mathematical model (reaction-transport) of this type of silicification was used to characterize the physicochemical conditions and to attempt a quantitative treatment of the phenomenon. Kinetics seem to be the limiting factor of quartz precipitation in the voids. However, the modeling shows that the kinetics of quartz precipitation limit the development of the silica replacement, whereas the diffusion of the dissolved species, from the replacement front towards the voids, seems to limit calcite dissolution. Limestone is replaced by silica, without any increase in the porosity, if the groundwater is close to equilibrium with calcite. The precipitation rate of the quartz depends on the number of quartz nuclei. The model predicts that silica deposition and calcite replacement can be completed in about 10,000 to 100,000 years.