Previous studies have shown several instances where airborne magnetic data, extrapolated to satellite altitude, differ from data obtained by the POGO and MAGSAT satellites. In an effort to explain this difference we have examined errors in the coefficients representing the Earth's gravitational field which could produce satellite positional variations and therefore would account for the observed differences. Our results indicate that while the magnitude of the orbital errors are in a range that could account for the amplitude of these observed variations, the frequency of the quasi-periodic orbital error is too large to explain the localized nature of the observed differences. Hence, uncertainties in the coefficients of the gravity model cannot explain the observed pheomneon. In a separate, but related computation, large crustal gravity anomalies have also been examined and shown to be insufficient to produce the desired magnitude of orbital perturation.