This document presents a new plate motion model, RM2, derived from an improved dataset of 110 spreading rates, 78 transform fault azimuths, and 142 earthquake slip vectors. RM2 represents a significant improvement over the previous model, RM1, with a mean averaging interval of less than 3 million years. The model performs well in most regions but has notable exceptions, such as misfits along the India-Antarctica and Pacific-India plate boundaries. These discrepancies are hypothesized to result from internal deformation within the Indian plate. RM2 also fails to satisfy the EW-trending transform fault azimuths observed in the FAMOUS area, which is attributed to closure constraints around the Azores triple junction. The model also struggles with the motion between North and South America, where the angular velocity vector remains poorly constrained. Absolute motion models derived from several kinematical hypotheses are compared with hotspot data, but the resolving power of the hotspot data is poor. The document discusses the revised data set, the general description of RM2, and detailed assessments of its performance along various plate boundaries. It also addresses the Indian plate problem, suggesting that internal deformation within the Indian plate may explain the discrepancies. The document concludes that the hypothesis of deformation within the Indian plate suffices to resolve the difficulties encountered in fitting the instantaneous relative motion data. The model's predictions and implications are discussed, including its usefulness for predicting global motions in areas where data are not available.This document presents a new plate motion model, RM2, derived from an improved dataset of 110 spreading rates, 78 transform fault azimuths, and 142 earthquake slip vectors. RM2 represents a significant improvement over the previous model, RM1, with a mean averaging interval of less than 3 million years. The model performs well in most regions but has notable exceptions, such as misfits along the India-Antarctica and Pacific-India plate boundaries. These discrepancies are hypothesized to result from internal deformation within the Indian plate. RM2 also fails to satisfy the EW-trending transform fault azimuths observed in the FAMOUS area, which is attributed to closure constraints around the Azores triple junction. The model also struggles with the motion between North and South America, where the angular velocity vector remains poorly constrained. Absolute motion models derived from several kinematical hypotheses are compared with hotspot data, but the resolving power of the hotspot data is poor. The document discusses the revised data set, the general description of RM2, and detailed assessments of its performance along various plate boundaries. It also addresses the Indian plate problem, suggesting that internal deformation within the Indian plate may explain the discrepancies. The document concludes that the hypothesis of deformation within the Indian plate suffices to resolve the difficulties encountered in fitting the instantaneous relative motion data. The model's predictions and implications are discussed, including its usefulness for predicting global motions in areas where data are not available.