Mechanical behaviour of a rock mass is significantly affected by the occurrence of inherent discontinuities (such as joints, fractures, bedding planes, etc.), therefore comprehensive knowledge of the effect of such discontinuities is necessary for the safe and economic design of geo-structures in or on a rock mass. Generally, the typical shear behaviour of a rock joint has been investigated under monotonic loading conditions, ignoring the fact that most of the time the structures embedded in a rock mass are subjected to cyclic loads, and it is likely that such structures fail to sustain the induced stresses from these cyclic loads (e.g., freight or passenger trains, seismic events or mining operations) because all the current standards for designing structures are based on analyses conducted under static loading conditions. The aim of this research was to experimentally investigate the effect of cyclic loading on the shear behaviour of natural rough rock joints under triaxial loading conditions and direct shear loading conditions, to semi-analytically model the shear behaviour of rock joints incorporating the influence of cyclic loading and to extend Sakurai’s (1977) concept of critical strain to the domain of rock joints considering the effect of cyclic loading. To achieve the aims of this study, several aspects of the shear behaviour of rock joints under cyclic triaxial and cyclic direct shear loading were studied. The experiments were conducted under cyclic triaxial and direct shear loading condition using GDS ELDYN triaxial apparatus and UOW servo-controlled direct shear equipment. Two different types of specimens were selected for both types of testing. The surface roughness characterization was performed using a non-contact 3D scanner (VIVID910) on each type of specimen, and the joint roughness coefficient (JRC) was evaluated as 12.6 and 7.2 for triaxial test specimens, and 11.7 and 8.1 for direct shear test specimens. For the cyclic triaxial tests, the test specimens were subjected to a cyclic deviatoric stress of 200 kPa and 300 kPa at a frequency of 8Hz with a varying number of loading cycles under confining pressures of 10 kPa, 20 kPa and 30 kPa. For the cyclic direct shear tests, the test specimens were sheared under constant normal load (CNL) and constant normal stiffness (CNS) boundary conditions at an initial normal stress of 0.25 MPa, 0.50 MPa, and 1.0 MPa, representing the typical insitu stress conditions observed in the field. The shear rate for the cyclic direct shear experiments was considered as