This study evaluates the passive mechanical response of porcine pulmonary arteries (n=6), focusing on the predictive robustness of the GOH constitutive model through multi-axial cross-validation. Proximal segments were excised into rings, longitudinal strips, square specimens, and cylindrical discs to perform opening angle, uniaxial (circumferential, longitudinal, and radial), biaxial, and confined compression tests. All procedures followed the 86/609/EEC Directive and were approved by the Ethical Committee of the University of Zaragoza. The analysis centres on the calibration and validation of the GOH model. While the model achieves satisfactory fits when applied to specific experimental datasets (error < 0.1), a significant lack of reciprocity is observed between loading modes: parameters calibrated from uniaxial data show limited accuracy in predicting biaxial outcomes (with errors reaching 0.6), and vice versa. The results suggest that incorporating radial experimental data helps to refine the model's three-dimensional stability, while confined compression tests further indicate that the tissue displays a reduced degree of compressibility. These findings highlight that integrating multi-axial data, rather than relying on a single loading state, improves the consistency of constitutive parameters for simulating the complex biomechanical environment of the pulmonary vasculature.