The shear modulus of the laminated glass beam structure plays an important role in modelling the structural behaviours such as lateral torsional buckling, inter-layer shift, bi-axial bending, etc. Determining the shear modulus of structures accurately is an extremely challenging task for academics and engineers. One commonly used method is torsion test as it provides a pure shear status. It is a relatively accurate way to obtain the shear modulus since it is to assess the torsional property of laminated glass beams by measuring the overall rotation rather than the component of the composite, eliminating the inaccuracy of other methods like analytic solutions to measure the torsional properties of the composite structure by an assumption to the property of the composite. However, many researchers are still utilizing an incorrect measuring method, calculating the shear modulus by torsion machine readings with the sensors mounted outside of the test specimen, to obtain the torsional property of structures. This study aims to identify a correct torsion test method that determines the torsional stiffness/ shear modulus of laminated glass beams with SentryGlas Plus (SGP) interlayer by conducting a comparative torsion test between calculating the shear modulus directly from torsion machine readings and obtaining it through the specimen’s rotation measured from the specimen’s central section using a tailor-made, non-destructive photogrammetry measuring technique. Additionally, eight reliable measuring devices, inclinometer sensors, were employed to validate and evaluate the measuring accuracy of the calibrated photogrammetric stereo-vision system. The torsion test of a monolithic glass beam sample was implemented to validate the feasibility of the torsion testing method for the laminated glass beam. Comparative test results suggest a large discrepancy in the shear modulus of the laminated glass beam between computed from torsion machine readings directly and measured from the central section of the specimen using the photogrammetric stereo-vision system and inclinometer sensors. A straightforward apparatus was used to confirm errors in calculating the shear modulus of the specimen from the torsion machine readings due to the rotation of the machine end, therefore the torsion machine readings should not be used to evaluate the shear modulus. This also raises the awareness of academics and engineers to adopt a proper and correct torsion test method for the evaluation of the torsional properties of structures, and then address the lateral torsional buckling problem in structural engineering effectively.