Rocks exhibit anisotropy due to their microscopic structure. Anisotropy of mechanical properties, especially the deformability, is an important parameter for geotechnical designs and overcoring stress measurement. Theoretical and experimental studies have been done to evaluate the anisotropy of rocks using elastic wave tests and loading tests. For soft rocks, the pressure dependency of anisotropy is considered not to be negligible. This paper deals with the method of evaluating the anisotropy and its pressure dependency of soft rocks. The method is based on loading tests instead of elastic wave tests, because anisotropy at the strain level of loading tests is suitable for the engineering purposes.Orthogonal anisotropy is considered here. The directions of anisotropic axes are treated unknown. A thorough evaluation of the anisotropy at low stress level has been found possible by uniaxial loading tests of specimens from 6 directions (X,Y,Z,XY,YZ,ZX). Four rosette gauges with 3 elements have been attached at the sides of each specimen. The direction of the main anisotropic plane analytically determined from the uniaxial tests coincided well with the bedding plane observed on the rock. Then hydrostatic pressure test has been adopted to measure the pressure dependency of the principal normal stiffness C 11 , C 22 and C 33 . The curves of these 3 stiffnesses and 3 parameters of anisotropic direction have been obtained against the pressure from 0 to 20 MPa.Thus an experimental and analytical procedure of determining orthogonal anisotropy and its pressure dependency has been proved. The rock used in this study is Opalinus clay. Some findings about this rock are that hypothesis of transverse isotropy may not be accurate enough even for a rock with clear lamina, and that C 11 /C 33 can be so large as more than 3 even at high confining pressure of about 20 MPa.
