AbstractA set of 48 cyclic and 12 monotonic large-scale direct shear tests was performed to assess the interface properties of sand–rubber composite along a nonwoven geotextile layer. Rubber content, semiamplitude of the shear displacement, and applied normal stress all were varied to determine the cyclic, postcyclic, and monotonic interface response of the composite system under shear loading. The test results show that adding 40% granulated rubber to pure sand caused approximately 50% reduction in the maximum mobilized interface shear stress as the loading cycles progressed. The addition of granulated rubber to the sand decreased both the damping and the shear stiffness of the interface for all values of displacement amplitude and normal stress; in particular, for the energy dissipation, the observations were associated with the higher linearity of the stress–strain relationship when adding rubber, thereby reversing the typical trend of higher damping at smaller strains or displacements. In addition, an increase in the displacement amplitude value yielded a reduction in the secant shear stiffness, but contrarily increased the damping ratio of the geotextile–composite soil interface. An increasing trend of the hardening factor was observed through the initial cycles of loading for the samples containing 40% granulated rubber, which was ascribed to the increased densification capability of the sand–rubber mixture with the progression of the loading cycles; however, this response was not captured for the pure sand–geotextile interface.