Repositories for high level radioactive waste hosted in argillite or granite with bentonite backfill are the two disposal concepts that been extensively studied. A general perception is that granite has very low sorption capacity and therefore is disadvantageous in retarding the migration of radionuclides in comparison with argillite. However, when there is bentonite backfill, does the sorption capacity of granite really matter for the migration of radionuclides? How different host rocks would affect the migration of U(VI) through the bentonite backfill of an engineered barrier system (EBS), and which properties of the host rocks cause those differences in migration, given that migration of radionuclides is a critical measure in assessing the performance of a repository? Here we present two coupled thermal, hydrological, and chemical models for the transport of U(VI), each with an identical setup except for the type of host rock (argillite vs granite rock). Comparisons between our models show that the different host rocks exert their influence on migration of U(VI) via regulating the chemical conditions in the bentonite, which affect the concentration of U(VI) at the source and the adsorption of U(VI) in the bentonite. The key chemical conditions are pH, Ca + , and HCO 3-concentration in pore-water, which are strongly affected by soluble carbonate minerals. Our models also show the occurrence of illitization (dissolution of smectite and precipitation of illite) in the bentonite, which affects the migration of U(VI) through changing pH and the quantity of adsorbents. Although generalization of current model results for granite and argillite should be done carefully, the simulations highlight the importance of pore-water chemistry in host rock, as well as bentonite-host rock interactions, when assessing the performance of a repository. In comparison to properties such as thermal conductivity, permeability, and adsorption capacity, properties that have been studied intensively, the pore-water chemistry of host rocks deserve equal or even more attention because of its strong influence on radionuclide migration.