The laser geodetic satellites Lageos and Starlette exhibit residual orbital motion with an unexplained seasonal component. In addition, recognized polar motion excitation sources do not account for a large portion of observed polar motion. It is hypothesized that air and ocean mass redistribution is the primary source of seasonal perturbations in satellite motion, and that wind-driven ocean mass redistribution is a major source for polar motion excitation. Average monthly variations in zonal spherical harmonic geopotential coefficients are estimated from NMC air pressure for 1958 through 1973, and from variations in continental water storage predicted by a global hydrologic model. These coefficients are used to predict average monthly perturbations in the longitude of the ascending node ([Omega]) for Lageos and Starlette, and in the eccentricity vector ([Psi]) for Starlette. WMO monthly air pressures and twice-daily Navy sea level pressures are used to predict time series of [Omega] and [Psi] perturbations for Lageos during 1976 through 1985, and for Starlette during 1980 through 1983. In addition, the Hellerman and Rosenstein wind stress field for world oceans and the Gill-Niiler bottom pressure equation are used to estimate annual and semi-annual ocean mass redistribution, and to predict polar motion excitation vectors and Lageos [Omega] perturbations. Comparison of predicted [Omega] and [Psi] perturbations with observed Lageos and Starlette behavior indicate that air pressure may be responsible for much of the unmodeled seasonal variation in the Earth's geopotential. In contrast, the water storage contribution is very small. Year-to-year variability in the observed Lageos and Starlette [Omega] times series is well matched by predicted perturbations. Even after the removal of annual and semi-annual components, significant coherence remains between predicted and observed [Omega] time series for both Lageos and Starlette at periods of less than one year. Comparison of predicted polar motion with ILS observations suggest that the effect of ocean mass redistribution is significant, and second only to air pressure in magnitude. Lageos [Omega] perturbations predicted from ocean mass redistribution indicate that non-isostatic sea level fluctuations should be readily observable by satellite laser ranging. / text