A series of complex experimental histories have been performed on two specimens of Nordland Shale from the cap rock of the Sleipner CO₂ injection site in the North Sea. By simultaneously applying a confining back pressure, specimens were isotropically consolidated and fully water saturated under realistic conditions of effective stress. Ingoing and outgoing fluxes were monitored at all times. Multistep consolidation and hydraulic tests were performed prior to gas injection to determine baseline hydraulic properties. Both specimens were found to be relatively compressible with a general trend of reducing compressibility with increasing effective stress. Hydraulic permeability, anisotropy ratio, and specific storage were quantified by inverse modeling using an axisymmetric two-dimensional finite element model. Estimates for elastic deformation parameters were derived from the analysis of consolidation transients. Both specimens yielded comparable intrinsic permeabilities of around 4 10¹⁹ m² (43 10¹⁹ ft²) perpendicular to bedding and 10¹⁸ m² parallel to it. Specific storage was found to vary with effective stress within the range of 2–6 10⁵ m¹ (0.6–1.8 10⁵ ft¹). Gas transport properties were determined by multistep constant pressure test stages, using nitrogen as the permeant. Analysis of the flux data indicates gas entry and breakthrough pressures under initially water-saturated conditions of 3.0 and 3.1 MPa, respectively. Using a stepped pressure history, flow rate through the specimen was varied to examine the underlying flow law and the possible effects of desaturation. With the injection pump stopped, gas pressure declined with time to a finite value, providing a measure of the apparent threshold capillary pressure, which ranged from 1.6 to 1.9 MPa. Numerical modeling of the gas data, using the TOUGH2 code, suggests that anisotropy to gas flow is greater than hydraulic flow. Fits to the pressure data were obtained, but matching the magnitude of the flux through the sample was not possible. Based on the data and subsequent model activities, standard concepts of viscocapillary (two-phase) flow are clearly inadequate to accurately describe the processes and mechanisms governing gas flow in the Nordland Shale. Evidence suggests that gas movement occurs through pressure-induced pathway flow, accompanied by a limited degree of viscocapillary displacement. The laboratory experiments support the time-lapse seismic observations that the cap rock is performing as an effective capillary seal. The experimental results also indicate that if gas flow is induced in this type of material, it is mainly via discrete pathways, instead of distributed Darcy flow. This is consistent with observed CO₂ flow patterns within the reservoir, although a satisfactory explanation for how such pathways develop remains elusive.