Key questions in fault reactivation in shales relate to the potential for enhanced fluid transport through previously low-permeability aseismic formations. Here we explore the behavior of a 20 m long N0-to-170°, 75-to-80°W fault in shales that is critically stressed under a strike-slip regime (σ1 = 4 ± 2 MPa, horizontal and N162° ± 15°E, σ2 = 3.8 ± 0.4 MPa and σ3 = 2.1 ± 1 MPa, respectively 7-8° inclined from vertical and horizontal and N72°). The fault was reactivated by fluid pressurization in a borehole using a straddle packer system isolating a 2.4 m long injection chamber oriented-subnormal to the fault surface at a depth of 250 m. A three-dimensional displacement sensor attached across the fault allowed monitoring fault movements, injection pressure and flow rate. Pressurization induced a hydraulic diffusivity increase from ~2 × 10-9 to ~103 m2 s-1 associated with a complex three-dimensional fault movement. The shear (x-, z-) and fault-normal (y-) components (Ux, Uy, and Uz) = (44.0 × 10-6 m, 10.5 × 10-6 m, and 20.0 × 10-6 m) are characterized by much larger shear displacements than the normal opening. Numerical analyses of the experiment show that the fault permeability evolution is controlled by the fault reactivation in shear related to Coulomb failure. The large additional fault hydraulic aperture for fluid flow is not reflected in the total normal displacement that showed a small partly contractile component. This suggests that complex dilatant effects estimated to occur in a plurimeter radius around the injection source affect the flow and slipping patch geometries during fault rupture, controlling the initial slow slip and the strong back slip of the fault following depressurization.