The Panthalassa Ocean, which surrounded the late Paleozoic-early Mesozoic Pangea supercontinent, was underlain by multiple tectonic plates that have since been lost to subduction. In this study, we develop an approach to reconstruct plate motions of this subducted lithosphere utilizing paleomagnetic data from accreted Ocean Plate Stratigraphy (OPS). We first establish the boundaries of the Panthalassa domain by using available Indo-Atlantic plate reconstructions and restorations of complex plate boundary deformation at circum-Panthalassa trenches. We reconstruct the Pacific Plate and its conjugates, the Farallon, Phoenix, and Izanagi plates, back to 190 Ma using marine magnetic anomaly records of the modern Pacific. Then, we present new and review published paleomagnetic data from OPS exposed in the accretionary complexes of Cedros Island (Mexico), the Santa Elena Peninsula (Costa Rica), the North Island of New Zealand, and Japan. These data provide paleolatitudinal plate motion components of the Farallon, Phoenix and Izanagi plates, and constrain the trajectories of these plates from their spreading ridges towards the trenches in which they subducted. For 83 to 150 Ma, we use two independent mantle frames to connect the Panthalassa plate system to the Indo-Atlantic plate system and test the feasibility of this approach with the paleomagnetic data. For times prior to 150 Ma, and as far back as Permian time, we reconstruct relative and absolute Panthalassa plate motions such that divergence is maintained between the Izanagi, Farallon and Phoenix plates, convergence is maintained with Pangean continental margins in Japan, Mexico and New Zealand, and paleomagnetic constraints are met. The reconstruction approach developed here enables data-based reconstruction of oceanic plates and plate boundaries in the absence of marine magnetic anomaly data or mantle reference frames, using Ocean Plate Stratigraphy, paleomagnetism, and constraints on the nature of circum-oceanic plate boundaries. Such an approach is a crucial next step towards quantitative reconstruction of the currently largely unknown tectonic evolution of the Earth's oceanic domains in deep geological time.