Different concepts are used for the analysis and transfer of mandibular movements to virtual or conventional articulating systems. Some common procedures and analyses include the determination of the terminal hinge axis. However, despite the widespread use of different methods for hinge axis determination, very little information on the applicability and quality of these methods is currently available in the literature. The aim of this study was to provide an overview of the methods already being applied and to search for novel algorithmic methods, comparing them with respect to achievable accuracy and implementation. This comparison was based on new extensive computer simulations, where the influence of measurement noise on the result of the hinge axis position could be investigated. The assumptions used for the simulations were set so that the conditions allowed for the most accurate hinge axis determination: this comprised a pure rotation during mouth opening, within an incisal pathway of 15 mm, a measurement accuracy of 50 µm, and an optimal positioning of the entire measurement setup. The results of the computer calculations show that the best accuracy can be guaranteed by the novel least squares method, introduced in this article for temporomandibular joint (TMJ) measurements. Additionally, only methods tracking two and more (iterative or parallel) independent markers or equivalent jaw position measurements provide enough information for reliable accuracy. Using actual technical equipment, the highest accuracies can be achieved in a TMJ-near measurement setup. However, even in that best-possible setup, the error of hinge axis determination cannot be expected to be less than ±1 mm. For a better characterization of actual electronic recording systems, manufacturers need to provide more insight into the evaluation processes. Keywords: terminal hinge axis, jaw tracking, temporomandibular joint, accuracy, computer simulation