Abstract The behavior of cementitious materials under severe loading is of particular interest, for research on ballistic impact, penetration and near distance explosions, where very high pressures are developed. Therefore it is important to investigate the behavior of such materials in the exceptionally high hydrostatic pressures range. One of the key characteristics of this behavior is the equation of state (EOS) that is the relationship between hydrostatic pressure and density (or bulk strain). However it has not been adequately investigated, and therefore the mechanisms of cementitious materials deformation and damage that are developed within that range of high pressures is at least partly obscure and far from being clearly understood. This is partly because the controlled application of extreme pressures requires special equipment and expensive experimental setups and testing is associated with a wide variety of technical problems. This paper aims at presentation of the development of an experimental setup to perform confined compression tests of mortar and concrete specimens at high pressures up to 1GPa. It presents the experimental study of different cement paste and mortar specimens and their comparison with previous results, obtained under high pressures up to 300 MPa. This allows the investigations of size effect for unsaturated samples, as well as the effect of the sand volumetric content on the loading branch of the equation of state. The experimental study shows a good repeatability for the cement paste specimens and for mortar specimens with fine sand. It also shows that decrease of water/cement ratio in the cement paste mix as well as increase of the sand volumetric contain in mortar results in monotonic increase of the secant bulk modulus of the loading branch of EOS. The comparison of the mean loading branch of the experimental EOS for unsaturated specimens having 30 mm diameter (70 mm height) and 70 mm diameter (150 mm height) clearly indicates that there is practically no size effect for all tested mixture compositions. While for the cement paste there is no difference up to 270 MPa (highest pressure level in the test in the small apparatus), the addition of sand yields an increase of the difference, which, anyway, remains small.