The prediction of large concrete structures behaviour such as bridges, dams and nuclear containment buildings (NCB) is a vital issue regarding the evaluation of the durability, safety and effect on the surrounding environment. In this work, experimental and numerical estimations of the saturation ratio profiles in the VeRCoRs NCB mock-up structure are presented, and their blind comparison is discussed. Non-destructive testing (NDT) based capacitive measurements are processed to quantify the saturation ratio experimentally. Then, a weakly coupled thermo-hydric (TH) finite element model is presented within the serviceability state of VeRCoRs for the numerical counterpart. The uncertainty in the capacitive measurements and the constitutive parameters of the numerical model are highlighted as coming from the following three sources: (a) the natural randomness of the material properties and physical processes, (b) the limited knowledge of some input parameters in the mathematical model either obtained experimentally at the specimen scale and (c) empirically derived in the exploitation of the physical process. The influence of the uncertainty in the saturation ratio estimations is presented by error bars experimentally and is addressed by a stochastic finite element (SFE) study in the numerical work. Accounting for uncertainties in NDT and SFE measurements, the mean values of saturation ratio (reported in %) profile are forecasted up to 40% and 34% respectively, with standard deviation for both in the range of 3.5–4.5% in magnitude.