Modelling of the material behaviour is crucial for machining simulations. Strain and strain rates can reach values of 1–10 and 10³–10⁶ s⁻¹ during the severe deformations associated with machining. An existing dislocation density model for AISI 316L based on a coupled set of evolution equations for dislocation density, mono vacancy concentration is enhanced in order to accommodate plastic deformation at high strain rates. Two mechanisms are evaluated with respect to their contribution in this respect. One is rate dependent cell formation and the other is dislocation drag due to phonons and electrons. Furthermore a temperature and strain rate dependent recovery and a proportionality interaction factor and short range component that both depends on the dislocation density are also considered. High strain rate compression tests are performed using Split-Hopkinson technique at various initial temperatures. Experimental results are then used to calibrate the models via an optimization procedure. Evaluation of various flow stress models shows that the flow stress behaviour of 316L stainless steel is best modelled by the model with a rate dependent cell formation. Its numerical solution is implemented in a format suitable for large-scale finite element simulations