As observed from the current literature, the vast majority of the constitutive models that have been proposed for the description of the changes in the high-temperature flow stress of high-Mn austenitic steels, in the course of plastic deformation as a function of deformation conditions, have been expressed in terms of the total applied strain to the material (ε), which, as indicated by Follansbee and Kocks, is not a valid state parameter. In order to overcome this disadvantage, which hinders the prediction of the changes in flow stress under transient deformation conditions, commonly found during industrial hot-working processes, a different constitutive description is proposed. The advanced approach is based on the temperature and strain rate description of three important parameters: yield stress, hypothetical saturation stress and actual steady-state stress, which is considered equal to the critical stress for the onset of dynamic recrystallization, as proposed by Jonas et al. This description is accomplished by means of the well established Sellars–Tegart–Garofalo model, employing a unique value of the apparent activation energy for hot-working determined from the analysis of the peak stress values, as proposed earlier by Glover and Sellars. The model also requires a description of the time for achieving 50% dynamic recrystallization, which is expressed as a function of deformation conditions by means of the simple parametric relationship also proposed by Jonas et al. This information is subsequently input into two evolution equations, which describe the change in the work-hardening and work-softening rates of the material, as a function of deformation conditions, either when the work-hardening is counterbalanced only by dynamic recovery or when it is counterbalanced by both dynamic recovery and dynamic recrystallization. The results reported in the present communication indicate that the proposed constitutive description is able to provide an accurate reproduction of the experimental flow stress values determined under constant deformation conditions, obtained by Wietbrock et al. However, the analysis of a hypothetical deformation schedule conducted under transient deformation conditions suggests that a second internal state variable, other than dislocation density, might be required for the improvement in the prediction of the corresponding stress–strain curves.