In recent years, there has been a push for the incorporation of robots into manufacturing processes. In general, parallel robots are preferred for processes requiring high repeatability and positioning accuracy. If the positioning accuracy of the end-effector of a robot has high priority, compliance characteristics of the elements of its mechanism should be considered. Due to the high accelerations or external loading on the robot, the dimensions of the elements change and this leads to positioning errors for the end-effector. This thesis describes an experimental test setup and an experimental procedure for determining the compliance characteristics of planar mechanisms, followed by a comparison of the repeatability and stiffness performance of a parallel and an over-constrained mechanism. Finally, assumptions and methodology for using this compliance information to improve the trajectory tracking accuracy of high-accelerated robots are given. Portable coordinate measurement machine and calibrated weights are used to collect compliance information. The compliance behavior of the mechanisms defined for entire workspace by using the least squares and bilinear interpolation techniques. The D’Alambert principle is used to estimate fictitious forces that cause the compliance of the mechanism’s end-effector while the mechanism operates at up to 5 g accelerations. As a result of this thesis, it is demonstrated that the mechanism’s center of gravity and joint types play an important role in the mechanism’s trajectory tracking accuracy, and that tracking accuracy can be improved by a simple data-driven compliance prediction algorithm.