It is usual practice in rheology to measure the yield stress in a simple shear flow. In these measurements, the yield stress is identified as the maximum value of the shear stress below which no irreversible flow occurs. Then, the thus determined yield stress is used in conjunction with the von Mises criterion in any complex flow. The latter compares it with the intensity of the deviatoric stress tensor. It happens that for simple shear flow the intensity of the deviatoric stress is composed of both the shear stress and the normal stress differences, but the contribution of the latter is never considered in the experimental determination of the yield stress. In view of assess the importance of the contribution of the normal stresses to the yield stress, a sequence o standard constant shear stress tests were performed for each material, estimating the critical stress which represents the mean value obtained between the stress values of the curves in which the material flows and does not flow with an accurate tolerance. After that, proposed tests were performed in order to obtain the values of N1 − N2 and solely N1 at the critical stresses. Following the appropriate yield stress evaluation. It was observed that for some materials the normal stress contribution is much larger than the shear stress contribution. Furthermore, the validity of the von Mises yielding criterion for elasto-viscoplastic materials was evaluated. For this purpose, in order to generalize the study for different flow conditions, constant volume squeeze flow and traction tests were performed evaluating the corresponding yield stresses. As the most important conclusion, the von Mises yielding criterion was considered not to be accurate representing yielding for the elastoviscoplastic materials analyzed.