The ability to predict track geometry degradation is of critical importance in planning of track maintenance operations. This paper presents results of an FRA sponsored study on the relationship between track geometry degradation and ballast condition as measured by Ground Penetrating Radar (GPR). The study examined six different sites on a major Class 1 freight Railroad, with a range of ballast conditions and tie types, and developed a relationship between the rate of degradation of key track geometry parameters, profile (surface), and cross-level, as a function of two GPR measured ballast parameters: Ballast Fouling Index and Fouling Depth Layer (depth of clean ballast layer). The study sites included both fouled and clean ballast conditions that were monitored for track geometry on a very frequent basis (every two to four weeks) and also experienced multiple GPR measurements during that same period. Data analytic techniques were applied to this large data set to develop a relationship between rate of geometry degradation and ballast fouling condition and depth of ballast. The results showed a statistically significant relationship between high rates of geometry degradation and poor subsurface conditions as defined by the GPR parameters. A predictive model was developed to project and forecast geometry degradation as a function of ballast conditions.