Abstract Purpose The aim of this study was to determine the accuracy in measuring the pelvic orientations of a phantom model using the PosturePrint method. Methods In the Université du Québec à Trois-Rivières biomechanics laboratory, Trois-Rivières, Quebec, Canada, a mannequin was fixed on a rotating platform. For a set of 3 photographs (left lateral, anterior to posterior, right lateral) of each position, the mannequin pelvis was placed in 68 different postures on a stand, 61 cm from a wall, in front of a digital camera. The camera was at 83.8 cm in height and at 3.35 m from a calibrated wall grid. Mannequin postures were in 5 degrees of freedom: lateral translation (Tx), lateral flexion (Rz), axial rotation (Ry), flexion-extension (Rx), and anterior-posterior translation (Tz). Average errors were the differences of the positioned postures to the PosturePrint computed values. Results Mean and SD of computational errors for rotation displacements were Rx = 0.5° ± 0.8°, Ry = 1.3° ± 0.8°, and Rz = 0.5° ± 0.3°, and for translation, Tz = 1.2 ± 0.6 mm and Tx = 0.9 ± 0.5 mm. Conclusions The PosturePrint system allowed for accurate postural measurement of rotations and translations of a mannequin pelvis. The next step in evaluation of this product would be a reliability study on human subjects.