The out–of–roundness of wheel treads are an inherent degradation characteristic of the wheels of railway vehicles, and the interface behaviour of the wheel and rail caused by out–of–roundness features has essential research value in high–speed train running safety and quality. Sliding–rolling contact wear testing was conducted to evaluate the effect of the out–of–round wheel on the wheel–rail damage behaviour, by using a twin–disc setup. The effects of four wheels, three wheels of different eccentricities and one normal wheel were taken as a control group. The characteristics of the wheel–rail interface adhesion, wheel surface damage and compressive residual stress (CRS) distribution were compared and analysed. The results showed a dramatic increase in the fluctuation amplitude of the wheel–rail vertical force of the eccentric wheels, and the adhesion coefficient of the wheel–rail interface significantly decreased. Furthermore, eccentric wheels dramatically accelerate the wear loss of the rail used as a counterpart during wheel–rail rolling contact. The surface damage behaviour of the eccentric wheels is obviously different along the circumferential direction, and non–uniform wear was observed. The damage mechanism of the worn surface presented varying characteristics, which could be divided into a severely damaged zone (located at the maximum of the rolling radius rmax) and a slightly damaged zone (located at the minimum of the rolling radius rmin). The surface contact stress of the severely damaged zone was high, this finding can be explained by the extremely rough worn surface caused by the severe fatigue delamination characteristics and the initiation and propagation of fatigue cracks in the severely damaged zone. Furthermore, the values of surface hardness and residual stress were relatively low. By contrary, an adhesion layer was observed on the worn surface of the slightly damaged zone. The adhesion layer protects worn surface in the slightly damaged zone was relatively flat, and the CRS was relatively high.