In finite element analysis of sheet metal forming the use of combined isotropic-kinematic hardening models is advisable to improve stamping simulation and springback prediction. This choice becomes compulsory to model recent materials such as high strength steels. Cyclic tests are strictly required to evaluate the parameters of these constitutive models. However, for sheet metal specimens, in case of simple axial tension-compression tests, buckling occurrence during compression represents a serious drawback. This is the reason why alternative set-ups have been devised. In this paper, two experimental arrangements (a cyclic laterally constrained tension-compression test and a three-point fully reversed bending test) are compared so as to point out the advantages and the disadvantages of their application in tuning the well-known Chaboche’s hardening model. In particular, for tension-compression tests, a new clamping device was specifically designed to inhibit compressive instability. Four high strength steel grades were tested: two dual phases (DP), one transformation induced plasticity (TRIP) and one high strength low alloy material (HSLA). Then, the Chaboche’s model was calibrated through inverse identification methods or by means of analytical expressions when possible. The proposed testing procedure proved to be successful in all investigated materials. The achieved constitutive parameters, obtained independently from the two experimental techniques, were found to be consistent. Their accuracy was also been assessed by applying the parameter set obtained from one test to simulate the other one, and vice versa. Clues on what method provides the better transferability are given.