Finite element modeling of the machining process has been quite successful in recent years. The model can be used to better understand the interactions between chip flow, heat generation, residual stress, tool stress, tool wear, tool chatter and dimensional accuracy. One of the key inputs to the model is the high strain rate (>104 1/sec) flow stress data. The split Hopkinson bar test has been commonly used to measure the flow stress at these high strain rates. The method is expensive and typically limited to the strain rate range from 102 to 104 1/sec. To overcome these limitations, an optimization based inverse methodology was developed. In this technique, the measured and predicted cutting forces were matched by iteratively adjusting the coefficients in the flow stress constitutive model. Since the method requires many FEM simulations to reach the final optimal condition, a computationally efficient FEM solver using ALE (Arbitrary Lagrangian Eulerian) method was adopted to make the method practical. The method was validated with experimental data with excellent agreement.