Relative motion between plates has been detected and quantified in a number of ways, particularly for the Cenozoic and Mesozoic. This project will extend this work with special attention to the accumulated errors involved in past plate position reconstructions. The use of covariance matrices to describe the errors in plate rotations appears useful and will be applied to all reconstructions performed. Plate reconstructions will be extended back to the Mesozoic Era, the earliest time for which significant seafloor remains. This longer time history of plate motions, and analysis of these, will provide a better opportunity to study the evolution of these motions and will improve our understanding of which features are characteristic of plate tectonics and which are transitory and accidents of a single geometry or time interval.