Abstract This study delves into the phenomenon of high-frequency squeal noise occurring as trains traverse small-radius curved tracks and investigates the factors influencing wheel–rail curve squeal noise, particularly focusing on stiffness matching. To achieve this, we initially construct a finite element model of the wheel–rail friction system using finite element software ABAQUS 2022, validating its accuracy against Coulomb’s friction law. Subsequently, we employ complex eigenvalue analysis to extract the complex eigenvalues and vibration modes of the wheel–rail system, enabling us to study the positions and vibrational patterns associated with squeal noise by analyzing the amplitudes of unstable modes. Finally, we assess the impact of wheel–rail stiffness matching on curve squeal noise, using wheel–rail material stiffness and rail support stiffness as key variables. The outcomes of this study reveal the following insights: (1) Unstable modes closely align with the resonant frequency and mode shape of the wheel and rail. (2) Curve squeal noise primarily emanates from vibrations at the rim, railhead, and rail foot. (3) Wheel and rail stiffness significantly affect squeal noise, with a significant deviation in the elastic modulus between rail and wheel increasing the likelihood of squeal noise, while an optimal ratio of about 1.2 is observed. (4) Rail support stiffness plays a discernible role in controlling curve squeal noise. Theoretically, maintaining an appropriate support stiffness level can minimize the negative damping ratio of unstable modes, providing a viable avenue for curve squeal noise control. Keywords: curve squeal noise; finite element; stiffness matching; complex mode