IntroductionOver many years, fine sediments resulting from weathering and other surface processes could subsequently ingress to rock joints, reducing the overall shear strength of the joint surface [1][2][3]. Rock joints that are naturally filled with fine materials (see 2) are likely to be the weakest elements in a rock mass and can have a dominant influence on its shear behaviour due to of the low frictional properties of the infill [4,5].The most effect of filling material is to separate the discontinuity walls and thereby reduce intact rock contact, but shear strength will also be influenced by the nature of the filling material itself and the characteristics of the wall-fill interfaces. Because of the lack of reliable and realistic theoretical or empirical relations and the difficulties in obtaining and testing representative samples, engineers generally rely on judgment, often considering the shear strength of the infill itself to be conservative. In critical cases, in situ tests may be carried out to provide site specific design criteria, but invariably amount of testing that can be undertaken precludes the establishment of fundamental relations. During the past 30 years much more information has become available on the shear behavior of joints infilled with soil material. Several models have been proposed to predict the shear strength of infilled joints under both constant normal load (CNL) and constant normal stiffness (CNS) boundary conditions, considering the ratio of infill thickness (t) to the height of the joint wall asperity (a), i.e., t/a ratio [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22]. The experimental researches to date have tended to focus on modelled joints or replicas rather than natural rock joints. In other words, although some works have been done on the effect of infill material by using natural rock joints, they have not proposed prediction models and it can be noted that most of the previous models in the literature have been developed based on the laboratory tests over the simulated artificial joints and not on the real rock joints (for simplicity and reproducibility reasons). In addition, there have been no shear behaviour models which consider differences in infill type within the rock joint. However, this research will give an account of application of a statistical analysis on a series of data obtained from a complete testing program on natural infilled rock joints, taking into account three different material fill-