Motion generation systems are becoming increasingly important in certain Virtual Reality (VR) applications, such as vehicle simulators. This paper deals with the analysis of the Inverse Kinematics (IK) and the reachable workspace of a three-degrees-of-freedom (3-DOF) parallel manipulator, proposing different transformations and optimizations in order to simplify its use with Motion Cueing Algorithms (MCA) for self-motion generation in VR simulators. The proposed analysis and improvements are performed on a 3-DOF heave-pitch-roll manipulator with rotational motors, commonly used for low-cost motion-based commercial simulators. The analysis has been empirically validated against a real 3-DOF parallel manipulator in our labs using an optical tracking system. The described approach can be applied to any kind of 3-DOF parallel manipulator, or even to 6-DOF parallel manipulators. Moreover, the analysis includes objective measures (safe zones) on the workspace volume that can provide a simple but efficient way of comparing the kinematic capabilities of different kinds of motion platforms for this particular application.