The increasing use of GPS to locate transportation data on highways, railroads, and other transportation features presents some interesting challenges to the dynamic segmentation data model. These include calibration of more accurate measures, three-dimensional referencing, and multiple lane representation. These challenges will be discussed together with possible extensions to the dynamic segmentation data model to accommodate the new paradigm of nonlinear GPS location. Examples will be used from recent GIS-T project experiences in several states including Texas and Arizona.