Although the Oberst Beam Test Method is widely used in practice, detailed information about how to perform a successful Oberst beam experiment is quite limited. In this paper, first, the effects of various parameters in an Oberst test rig, including the amplitude of the excitation, mounting conditions, input excitation type and the length of the test sample, are examined in an attempt to improve the accuracy of the estimated material properties. As it is observed that the electromagnetic effect created by a non-contact exciter can be the most significant source of error in estimated material properties, this paper then presents the results of extensive tests so as to quantify the level of the adverse effects of non-contact electromagnetic excitation. It is found that non-contact electromagnetic exciter creates a stiffness effect that can be modelled as a spring attached between the non-contact exciter location on the Oberst beam and the ground. In contrast to the common belief that the use of non-contact electromagnetic excitation has little drawbacks, it is shown that such excitation can introduce very significant level of errors in identified material properties. This paper also proposes a method for removing the adverse effects of the electromagnetic excitation in order to obtain more accurate material properties for uniform as well as composite beams. (C) 2012 Elsevier Ltd. All rights reserved.