In metal forming operations the stress and strain levels can locally reach much higher magnitudes than those measurable in a standardised uniaxial tension test. Additionally, the stress and strain states are in many cases multi-axial. In this paper an inverse method to obtain material data is proposed. The aim is to yield more accurate data for a wider range of strain compared to a standard uniaxial tensile test. The outline of the work is that a forming experiment is designed to reproduce tensile strains present in a full-scale cold forming process. The blanks used are made of relatively thick high strength hot-rolled steel. Process data from experiments, i.e. punch force and punch displacement, are used as input to an in-house optimisation software package. The direct problem solved in the inverse modelling and optimisation scheme is a finite element analysis (FEA) of the experiment. The goal is to find parameters in a constitutive model of the material that minimises the difference between experimental and FE-calculated data. The experiments are modelled in a commercial FE software. Four different isotropic hardening laws are used in the FE-model. One of the optimised models is applied in a forming simulation and geometric optimisation of a demonstrator part.