Applying Cenozoic Plate Motion Models to Proterozoic and Paleozoic What drives the plates? Do they drive themselves by sub- sidence and subduction of cooling lithosphere, or are they pas- sively carried by mantle convection? The most direct evidence is kinematics, the plate motions themselves. Present motions are over-determined and suggest slab-pull dominates other forces. Stress observations give additional detailed information about the forces driving plates, in that sets of forces producing the same net motion can have quite different stress patterns. Plate reconstructions for the past provide the only available test to determine which features are truly characteristic of plate tec- tonics on the earth. Analysis of global plate motions shows that many characteristics of current plate motions have persisted over the Cenozoic Era. Using these reconstructions to estimate plate forces for the assumption of a dynamical balance between active forces (slab-pull and ridge-push) and plate drag, as the passive torque, yields fairly stable forces over the Cenozoic, with the misfit increasing systematically for earlier reconstructions. Extending analysis back into the Paleozoic and earlier is diffi- cult because no magnetic anomaly data remain in the ocean basins, so data from a variety of sources must be evaluated and integrated to produce reconstructions. Over the last 600 m.y. plate speeds show considerable variation. The opening and clos- ing of the ocean between Laurentia and Gondwana 560-400 Ma pro- vides a test of dynamical models, and only certain ranges of parameters can model both the opening and subsequent closing. In the Proterozoic the main support for proposed supercontinents rests on apparent polar wander paths of the continents; however, these are both sparse and contain an inherent uncertainty in polarity of the field at any given time. Consistency of pole paths with uncertainties in age and pole location tests the existence of early supercontinents such as Rodinia.