Uniaxial and cruciform biaxial tensile tests are performed on a face-centered cubic material AA6016-T4 and a body-centered cubic material DP490 under 19 different loading paths, i.e., uniaxial tension in seven directions, simple shear along the rolling direction, and biaxial tension in rolling/transverse and 45°/135° directions, with seven and four stress ratios, respectively. The ability of several yield criteria to describe the plastic anisotropy for the tested materials, under the associated and non-associated flow rules, is evaluated systematically. A new evaluation strategy to check the validity and applicability of the material models is proposed in this study. This strategy is different from the traditional evaluation strategy in that the biaxial tensile mechanical properties measured in the 45°/135° sampling direction are included in the investigation scope, and it is not limited to the rolling/transverse sampling direction, i.e., the yield stress functions are evaluated by comparing the experimental and predicted uniaxial yield stresses and plastic work contours on the normal and diagonal planes. Further, the plastic potential functions are confirmed by the description accuracy of uniaxial rα-values and the directions of plastic strain rate in the rolling/transverse and 45°/135° directions. A new analytical calibration program is developed for the material parameters of the Poly6 yield criterion related to normal (a1,a2,...,a7) and shear stress components (a8,a9,...,a16), which can further introduce the yield stresses under the near plane strain states (σPS0, σPS45, and σPS90) and pure shear yield stress along the rolling direction (τ0). The results indicate that the new calibration strategy of Analytical Poly6-II&2 can accurately describe the anisotropic yield behavior of the tested materials compared with other advanced yield criteria, especially for plastic work contours on the diagonal plane. Significant anisotropic yield behavior was observed under biaxial tensile stress states with the same loading ratio in different sampling directions. Therefore, investigating the prediction accuracy of the biaxial tensile mechanical properties in only the rolling/transverse sampling direction is insufficient for evaluating the effectiveness of the material models; the mechanical properties at the initial yield point cannot fully reflect the plastic anisotropy of the materials. The material parameters of the Analytical Poly6-II&2 yield criterion are expressed as functions of equivalent plastic strain, which can continuously capture the evolution of the anisotropic yield behavior for AA6016-T4 and DP490. Three different calibration strategies of the material parameter a16 are evaluated in this study. These strategies do not affect the description ability of the uniaxial yield stresses, rα-values, and plastic work contours on the normal plane; however, they influence the calculation results of plastic work contours on the diagonal plane. For attaining a balanced prediction accuracy, it is recommended to provide the values of both τ0 and σPS45.