The acoustoelastic effect links elastic wave propagation to the constitutive and microstructural state of solids. While classical formulations adequately capture the effect of elastic deformation [1] and monotonic plasticity dominated by isotropic hardening [2], cyclic deformation involving kinematic and combined hardening mechanisms is not consistently represented. This contribution presents a computational framework for modelling the acoustoelastic response of metals under cyclic plastic deformation. The macroscopic material behaviour is described by a von Mises plasticity model with combined hardening of Chaboche type. At the micromechanical level, the evolution of the dislocation structure is represented by a Kocks–Mecking-type model [3] that distinguishes between mobile and immobile dislocation populations. Mobile dislocations are associated with the evolution of the backstress and therefore govern the kinematic hardening component, whereas immobile dislocations contribute to the isotropic hardening through an increase in the flow stress. The total dislocation density evolves according to a balance between dislocation multiplication and dynamic recovery, resulting in a bounded, exponentially saturating evolution towards a stabilised state under cyclic loading. The dislocation density is coupled to a generalised Granato–Lücke-type formulation to compute the relative change in ultrasonic wave velocity. The resulting acoustoelastic response exhibits a nonlinear evolution with progressive saturation, reflecting the stabilisation of the underlying microstructure. Numerical simulations for cyclic strain histories demonstrate that stabilisation of the ultrasonic velocity coincides with a saturation of the dislocation density and the cyclic stress–strain response. Kinematic and combined hardening primarily affect the rate of convergence and the limiting value of the acoustoelastic response. The proposed formulation provides a consistent basis for computational analyses of wave propagation in cyclically deforming solids