Abstract: For the same deep mudstone, three testing methods of permeability (conventional testing method, variable confining pressure method, and pulse method under triaxial compression) have been conducted respectively. According to the testing results, large difference of permeability was found inbetween the three methods, in which the permeability of conventional method is 3-5 orders higher than those of the other two methods. The results of variable confining pressures method while its confining pressure exceeds about 5 MPa approach relatively the results of pulse method under triaxial compression. Permeability of pulse method slowly decreased with the increasing of deviatoric pressure, and dilatancy didn’t occur even when confining pressure reached 30 MPa. The results indicate that the permeability of mudstone is strongly sensitive to pressure; the intrinsic reason of which is that the cracks, fissures and pores in rock matrix tend to be compacted under high compression, which results in the reduction of pore size and connectivity for percolation channels and then causes the permeability to decrease; moreover, samples production, oven drying treatment, etc. also influence some of the permeability. Based on the results and the in-situ stresses conditions of mudstone, it is proposed that the variable confining method is suitable for the permeability and its development of reservoirs’ cap rock, while the results of pulse method under triaxial compression is more suitable for the cap rock of energy storage caverns and rock wall of wastes disposals. However, one should be very cautious while uses the results from conventional method. The research carried out supplies some constructive references for the analysis of the percolation properties and sealability of mudstone in deep strata.

Key words: mudstone, permeability, confining pressure, pressure sensibility, deviatoric stress