It is important to evaluate the porewater chemistry of argillaceous host rocks for safety assessment because it affects radionuclide migration. The nanometric pore-space architecture and the resulting very low hydraulic conductivity, however, make the sampling and characterisation of porewater chemistry a difficult task. Thus, in order to understand and reduce uncertainties induced by sampling and experimental disturbances, a multi-method strategy is generally required. Here a combined experimental and modelling approach is presented for the analysis of the porewater chemistry in a rock sequence comprising Opalinus Clay and so-called " confining units" 'Brown Dogger' and Staffelegg Fm. (Liassic). The data was obtained from analyses on drillcores of a deep borehole at Schlattingen (NE Switzerland). The studied rock sequence can be roughly divided into four units: The top unit (∼25 m thick) corresponds to the lower part of the Effingen Member (Malm) and is made up of silty to sandy calcareous marls. The underlying unit (∼77 m thick), termed 'Brown Dogger', exhibits rather variable lithology. It consists predominantly of clayrocks and marls with variable contents of calcite and quartz. The third unit is the ∼120 m thick Opalinus Clay, which is characterized by fairly homogeneous silty to fine sandy clayrocks. The lowest unit (Staffelegg Fm., ∼53 m thick) is also rather rich in clay minerals and consists of a variable sequence of marls, clayrocks and few limestone beds. Only the three lower units were considered for porewater chemistry modelling. Experimental: Rock core samples representative for the overall lithology of the corresponding stratigraphic unit were taken about every 5 m, sealed on-site according to a standardized protocol to protect them from oxidation and evaporation, stored cool and sent to the laboratory for subsequent analyses. Mineralogical compositions, water contents and density measurements were obtained for 30 samples. From water contents and grain density, water-loss porosity was derived. Aqueous leachates were prepared at a solid/liquid (S/L) ratio of 1, with contact times of 48 h. Cation exchange properties were studied using the Ni eythylenediamine method at S/L = 1. All procedures, except for milling, were carried out under N 2 atmosphere in a glovebox. Squeezing tests were conducted on 12 selected clay-rich samples. Porewaters were extracted at 200 to 500 MPa, but only data from the lowest squeezing pressure of 200 MPa are reported here. Porewater was also extracted with the advective displacement technique from a core sample of the 'Brown Dogger' unit. CO 2 partial pressure measurements were carried out on well-preserved cores in gas cells. Model setup: The aim of the modelling was to simulate porewater compositions using primarily chloride data from leachates and exchangeable cation data and to compare these with the data from squeezing and advective displacement. Leachate chloride data was recalculated to in-situ porosity by considering water contents and a uniform anion-accessible porosity of 0.5. The latter was based on the comparison of leachate and squeezing data (Wersin et al. 2012). Furthermore, pCO 2 as obtained from pCO 2 measurements was considered as input. The applied thermodynamic model was based on the studies of Pearson et al. (2011) and Gaucher et al. (2009). A base case and several variant cases were defined, all of which assumed equilibrium with calcite, omnipresent in the rock sequence. Further, sulphate equilibrium with respect to celestite or a CaSO 4 phase was considered. The uncertainty in pCO 2 was addressed by variation in pCO 2 in a specific model case.