Many techniques have been developed for the purpose of analyzing and evaluating mechanical systems for the purpose of improving design and control. Such work is not necessarily applicable to mobile rovers, since the wheel–ground contact, which does not exist for fixed-base systems, plays a critical role in the behaviour. This thesis investigates important concepts related to analyzing rovers. In initial investigations of existing methods, it was found that current numerical simulation tools can efficiently model rovers, but are often limited to forward dynamics simulations, and are not suitable for more in-depth analyses. Additional investigations were performed to determine which contact models could be used, and it was found that wheel–ground behaviour was best captured using terramechanics-based models. Several concepts relating to defining mobility were investigated, and two new measures were proposed (maximum net traction force, and accelerability), which were found to provide useful, real-time indications of a rover’s capability. A novel method was proposed for analyzing the mass matrix of a mechanical system and four concepts were presented that provide information relating to the effective masses and dynamic coupling of a mechanism: locked effective inertia, force coupling, free effective inertia, and acceleration coupling. Finally, the operational space formulation, an important concept primarily used in the field of manipulators, was investigated and extended to be used with rovers. It was found to be a useful method of evaluating and controlling rovers with the appropriate selection of the operational space representation.